Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
docs: correct product site capability claims (#3500)
zhifu gaod7178079e033
diff --git a/web-pages/product-site/content/legacy-manifest.json b/web-pages/product-site/content/legacy-manifest.jsonindex fc07e3b86..2e9bdc012 100644--- a/web-pages/product-site/content/legacy-manifest.json+++ b/web-pages/product-site/content/legacy-manifest.json@@ -7,7 +7,7 @@"blog/cantonese-speech-recognition.html": "f4a674f2b8f15963a386d586bb79d62f9f75c20d704ca2eafba21a800f52d4df","blog/chinese-speech-recognition.html": "a78eeb4e5ea38ce95ad05ebc54eceb0185d8dddd2286769b5ba1b7dd00e8b615","blog/fun-asr-nano-guide.html": "4e9a5c41c3846e482cc74e5379e642b16b6b35b680eea36a4cc79147933b76d5",- "blog/funasr-cli-transcribe-command-line.html": "84dd389969e2795bf8bb9054b2f88b2abc16893bf6f7a7673e1f2887827da12e",+ "blog/funasr-cli-transcribe-command-line.html": "ebb7bcdb8b2a2a492dfb9b44844e11626e2b3e3469f48a18517f87182d65b852","blog/funasr-llama-cpp-whisper-cpp-alternative.html": "3e1a8901c3f7f74e310b8050e2f6722c736ce0b37155805b85786d5b5455ae8f","blog/funasr-realtime-streaming-asr.html": "7b6c34ace6c4c782f5af9c3f52181c1b1ef08d32c4ea3fafab63b1f30a9e72db","blog/funasr-speaker-diarization.html": "451273993769f38de014160788581529788db8e4ca9fb56773d7723c51d96a52",@@ -19,17 +19,17 @@"blog/funasr-vs-faster-whisper-chinese.html": "bfe9bb8017be80c7e7f4587726f43f6064f1dc4c65e39788f836fdfb0c9789f7","blog/funasr-vs-whisper-benchmark.html": "b7b49adf24d20570abb09b733ce03d4a50a4a0e98e746b4a9320f453e01cce84","blog/funclip-v2-1-0-video-clipping-release.html": "88f6c44e5332d1746c4db0fc97d755ef82f12e46d8f1c1c0ad9351152ff9dfc1",- "blog/generate-subtitles-srt-vtt-from-audio-video.html": "dbaf77072a980c8c50f33cb74e9513dea5dba2d31bd802fade38dcb0bb97e849",- "blog/index.html": "c7b584195fbb15ae797092e4b3305ee811822b6e86223e016f96c5497d0a4e8e",- "blog/japanese-speech-recognition.html": "90f1f3bee4dcd72e414c6cb1ef9c0e580d5b7107356bc4285ac801affbb64486",+ "blog/generate-subtitles-srt-vtt-from-audio-video.html": "f1133235673d441cc654310581f59a319f182314ab7c0b0824af87da9f4a0591",+ "blog/index.html": "41627afff22279325c3d5344514ed1fb689c1010f4f139ffb2a44bab3d4c2cf2",+ "blog/japanese-speech-recognition.html": "399f5ce84e68ac00854bdf70b52b1cde6795efca7c800fd6492035a37a7c1b68","blog/lightweight-speech-recognition-cpu.html": "d6270066222ed5baef0df108e6a4155f4a87bb5482169f23f218794d371f3281","blog/punctuation-restoration-python.html": "6cd26e03bf75b4343afd7b351c2681ebd513b058fb549c5888802be9ff3a1229","blog/self-hosted-deepgram-assemblyai-alternative.html": "13b4287b5ef0d3159bf0af09ece3bc1fea66f13390b13c7f33b8b5e85351a1f5","blog/self-hosted-google-aws-azure-speech-to-text.html": "8e4000fb9e2b33a48a3907d4bd0f55db57ba74121a89b5fa12b412c3c963f49a",- "blog/self-hosted-openai-whisper-api-alternative.html": "a749ad220d60398f8b1a637cacf68c48896a745350e1e982249caf32275c5002",- "blog/sensevoice-deployment-guide.html": "b2b05abdab33ceb40c9baefc24eb19c82068eb8fc1e87e76aa15d565bccfd28e",- "blog/sensevoice-emotion-language-detection.html": "6cd2a98e7813921754d33af9b2898f83dc9b1139482a34f2d0bb665189da67ee",- "blog/speech-to-text-python-transcribe-audio.html": "273c5e6fc6581ae995489d59c9f011de93a6c15e31f08fff56595bc0e6fd866f",+ "blog/self-hosted-openai-whisper-api-alternative.html": "8e7352392dc755428bd272c208107a64d9c6d5d1700ef52b42553a3ec234e3fd",+ "blog/sensevoice-deployment-guide.html": "61c34fb7626444a0f1b5c93013dd95c10c576333791eaca1ab819baf0a8cdad6",+ "blog/sensevoice-emotion-language-detection.html": "a00c40bfbaa3f0f390bc00f052b7e4e299a7d193b088014b7515813f1f8a636b",+ "blog/speech-to-text-python-transcribe-audio.html": "87335710235dfdec9e01288f03a9a294c7bc3f730110e8d06f2b4c9d4906e284","blog/speech-to-text-timestamps-python.html": "78364db117dfbce45b03739679764cec7ab1d0a860f321c67154f33392fc6b8d","blog/subtitle-edit-fun-asr-sensevoice-local-subtitles.html": "f8d5870b6cf9dbdfd4e943fdccb7ae8f091472f1dcbd46bd7db90bd36e3e7846","blog/voice-activity-detection-python.html": "9055206633caecb83c11031f273c8c9297852f44dfe50984e03ab59a7deda6b5",@@ -39,11 +39,11 @@"css/index.1682178c.css": "b5f6034ce886c25773928c114e4290d1c491a3d081677b488810198efd76b4e1","decoder.js": "f2023a28036b1ff58e5bdabbb72ea427f8faaf60e89acd78ff183c313bdb0a2c","donors.html": "eec6b789b29c27d372925ae0ad3e13d4c6e62978b3c38c4da9e3aec129f10934",- "ecosystem.html": "1007c2bd0cd6fffb849aa8aa651c077f9c4f52d0c171b8ffb1137c47c4bfe9e4",+ "ecosystem.html": "35c5507fc3cabb519aaa657208e187c44e323f3115befab11af262afa836abbf","en/blog/cantonese-speech-recognition.html": "b9bd146198206f781c37f341c370a690dc83cd7163334e34b4e7f65e69af8ff2","en/blog/chinese-speech-recognition.html": "16b68bd2a1122236c580c62275002ae2c0875ca14e32686e88d3b5ccfa9f56ef","en/blog/fun-asr-nano-guide.html": "1746c708bd0606188de4266d8255984dd9e14634af33c54ad65a5f05b767be1b",- "en/blog/funasr-cli-transcribe-command-line.html": "1675699115f99e629d59f2a5323389d9405a6044c2a54efbcdd682c883c9322b",+ "en/blog/funasr-cli-transcribe-command-line.html": "f174bac1351185aeacee3d1a885956dd7fa78bdc1992afb26de43e93bcf11f86","en/blog/funasr-llama-cpp-whisper-cpp-alternative.html": "3808ee7a014f63e3befd83c886eca3fd96abea7d4de68a90a8bd23befe579282","en/blog/funasr-realtime-streaming-asr.html": "65c7353fcdc393fc0adec3969d9762edc8f30dc15d20e360a2b11508d8c60a91","en/blog/funasr-speaker-diarization.html": "db6511de2f356b7bf5e3f1f89b33991500a293c9d3bd2420d8a6cd6cf6cf3997",@@ -55,28 +55,28 @@"en/blog/funasr-vs-faster-whisper-chinese.html": "abf31d3827dfba9b31ccd4e76229de9e4bfff00ab8bcee33bd718c88249630f4","en/blog/funasr-vs-whisper-benchmark.html": "367d4a8a1cc09ac932c80cdad065127925c5e5f942a806ace683c16dc1132769","en/blog/funclip-v2-1-0-video-clipping-release.html": "229baf59adf2c3290541d9b3c8a6243992406ba84b712714e9d14599205cb1d4",- "en/blog/generate-subtitles-srt-vtt-from-audio-video.html": "bd81aa43c6a977f28dd3925861bf456002fbacfec0df0504b413daa64ae60dfa",- "en/blog/index.html": "ac522ccc02a3df9688fa3ea21f13480f2ca4c996b7dcd4f66fbb28806775ccbe",- "en/blog/japanese-speech-recognition.html": "8b106563142c291d3b86fad79a51e9c8746417b7750430e91296cc90764e570f",+ "en/blog/generate-subtitles-srt-vtt-from-audio-video.html": "1530d4b9e94820a0b60d801e8d1c19b7aeb031b2f2b76c092657392e47706c6d",+ "en/blog/index.html": "828bcfa76d0911e13340fd03bdbea8a97c6a3407d04914b4d72564da25214cec",+ "en/blog/japanese-speech-recognition.html": "c476adcc2be1ed7c19e2345cc91b8ee6a0e04efd794b9b6476dd5d9b3a2c8b04","en/blog/lightweight-speech-recognition-cpu.html": "0e02c0e0853b12613d32c250f593c05c2052f18231207b63ee01d6e6600a86f7","en/blog/punctuation-restoration-python.html": "d6e30049f48d9dc6e6bc22eb567830013e4161eca51215616738b675dd1ffc40","en/blog/self-hosted-deepgram-assemblyai-alternative.html": "73b2fc101101fd2408ed49fdb2b0995ada59dec0016a91ed49efb856ebb8de53","en/blog/self-hosted-google-aws-azure-speech-to-text.html": "4a8453451722b3ec8735224a586744ddfbfec9458cedd433f9fc908351027e9a",- "en/blog/self-hosted-openai-whisper-api-alternative.html": "d6b4e0cfb26dd535ba634f01f1cc5a2e5cc2d71ce7d8e745698b33f13e7ca9cd",- "en/blog/sensevoice-deployment-guide.html": "c8e06e3ff6bd3ba333165faf047715107cccdb9e9df15bc52ca48de3d427cae7",- "en/blog/sensevoice-emotion-language-detection.html": "35ea4c9fc89b95b7b0de75c6138058717271d22222d5d0c502add275c068c794",- "en/blog/speech-to-text-python-transcribe-audio.html": "225d4334c591d53aa126d4f31bc4eb0cbaf0afb63c2486c7c934757a0a3627b9",+ "en/blog/self-hosted-openai-whisper-api-alternative.html": "cda3b8e45857996031b888b2bf076666b7c3dcae355f097e942fd770a3f0b729",+ "en/blog/sensevoice-deployment-guide.html": "8882912c6a84ec0a0e025d0eff8fcfa74bf76324bd4c9cae862c43c87c6e10b3",+ "en/blog/sensevoice-emotion-language-detection.html": "4fa0d49d34df3b923323fc49fbb02c432efd7cd0d90d81e84621dbbecc08cd29",+ "en/blog/speech-to-text-python-transcribe-audio.html": "4fc57aef3c6177e28afa26fa6c41f67280b986e384b32a63a2f79ca67bcfe631","en/blog/speech-to-text-timestamps-python.html": "6a7f8368f757c9ea50e287db6db3191ebcbbfc914f9a869b9515648b2d69790b","en/blog/subtitle-edit-fun-asr-sensevoice-local-subtitles.html": "7c65a73bab1ff481d9b398476ea6440926546ba7122cc2cb64f5bda81fc039a2","en/blog/voice-activity-detection-python.html": "d12f5d0f4ae07c3434f6188d12f5603b8abb9578413b79133668638d1c5e18a7","en/blog/which-funasr-model.html": "1207180aa929a3aa51db136c1098f4edc8f07405e2b56fe033998bf996f8ec7f","en/donors.html": "eacb7f9434b1b29e13e9b0175bb9680556ee72ab4dbd155e68f9c912eb0212b1",- "en/ecosystem.html": "c3e4f9e007953799a090565af0a7bcceaf66ae645c6d42591675080fb2d1a7dc",- "en/index.html": "4950bc2e0b337637300c5bcd45d1820d9636c4f1dab34e0ec8d3443be44bd277",+ "en/ecosystem.html": "93e688e4c7274a06e3bae0c9aef60251419fa1bb7c72352ea74607fde69c883d",+ "en/index.html": "d86effc3bf4d218eabd465cbbe026f62c98dd50110ddd58251d9d271a64882a2","en/llama-cpp.html": "dbfbd125ebbf07f7d923c91b122818f8b09484ce54c1e80eba66a89d22dfb5b4",- "en/models.html": "c2add621edc3aa63e6d49587f089ee5ac6c05faf5f68e14fef70a7a5ab443a4d",- "en/quickstart.html": "2735cc9e608bfc7860eaa0c0e1242088d98204bca92cfec7ae7923a056f968c4",- "en/vs-whisper.html": "1779f98b2eae8579f3694577bc9dda81bd954423b49dd201ff17dcfb5e630ab1",+ "en/models.html": "321a63caace107d5ce6d470948c2f5954f2d8e25306d8227b4a85ec3663ad49e",+ "en/quickstart.html": "7453cd7d582079eb4082373bce6e5de26492458b73c41617bbb1bd4248bda6ac",+ "en/vs-whisper.html": "4833b93195957210436f791a9d8ebc726f0af93b809d740a5cee1e53de0ac390","favicon.ico": "e5edd9f12088fa920b848485c62c72a093f27bfba7b6a46b4e7b63c34eb5126a","img/banner.4f436d19.png": "d4fbab32267455f6b54f9af414d8e9847ab13812ef1f89f39a71f485bc0e0210","img/funclip-v2-1-0-interface.jpg": "35855be22612faa4e6ed8900a1dea33d793b33e7b1c2a32a768718d806c6ff69",@@ -87,7 +87,7 @@"img/mxjs.6724be62.png": "9d398c1aa1e55de7b3d7729886d5f2866c31f00e82c3b36ea7740d02e60ee1aa","img/sstx-bg.d7e580c1.png": "33a6bd3159f3748244f3c27545314010a33bb7d9749c95be3fc24f0075741a62","img/subtitle-edit-speech-to-text.png": "3ea166b3dc21ab262418d40616852360e1c10aaba22a2365bfd4b1c238972e15",- "index.html": "95aa61940d7e48d3210899ec4b7d22bbd198ef7e3721b117e1e489f1e7b47b1a",+ "index.html": "fbd3edb1a4497dd8a35e0e6a5e05b16bed098cdaf7e879d2ba6d9eafef063090","index.js": "f10e5fff91ed36af5377609eefa067c7e22278d0dd8e558f065733ca99b93fd6","jessibuca.js": "02240fc0fdfa74ebe840cc6c818fc91a9c3cb830d9ebeddd9ca21b46168aa06f","js/742-legacy.a8705541.js": "bf11ce6893c6bb36593ee22ca053c94808f8d303925e1ab7366062bc467d1597",@@ -99,8 +99,8 @@"lang-redirect.js": "e1189080a8e81eb76cf39e0538d0378b6631f5aa87e4679c0fd4fd34d1f6c197","llama-cpp.html": "892b5c05e3375bcc7cd0af343b629f069d8ec4f372e6470cd830cdf2181a11d5","logo.png": "ac0ff97f2b0203c3fad4ac5e90231d93e36d98cc8e90d383483de150b0dd0728",- "models.html": "2fa557c327c81717e5ae145ae96436168ea51f58301bd7bb6357d4e15686cdfc",- "quickstart.html": "90c5f25b428b064be02df11c38d24556ec97f4bc80900398c223424aaa48e7b8",+ "models.html": "724c88917c9934ffa66948cf3eb583fed930dd689d691b7d21fa22ee8332e606",+ "quickstart.html": "a543bf07c2cfc9a69ef3831f15c9e9d3e571a826a73cf7e4bdc8e9eb7b910254","robots.txt": "1305b0f680b3e9f73b01301dc474423230dffab85731a379a694f30c73c07f2d","sitemap.xml": "967778b911c834ffeb186126353b4e8cecc009c5c73d4684b9978363e97ede34","static/liveplayer/liveplayer-component.min.js": "ea10fda35574cc3a77a3e7187095467d1a409009a2d39690051d33d0e7055d30",@@ -118,6 +118,6 @@"static/online/wav.js": "63344c3e282b80349bdc64f7508f32346ce783609881dbfbc97827a36d594c35","static/online/wsconnecter.js": "a619a1cab5e111c5ff4b77ec8e68e47f9dde6ba5d7a62ca2db0c1b083c33e8ca","voice/index.html": "1a6bae9e1637fea91f2f97506f5847ebd0ce429136444afa61f0a75da0beea0b",- "vs-whisper.html": "f465fc0cf17d4fee209c23b621058f3e87d816ef77402a18b509aa32b56864e3"+ "vs-whisper.html": "7d73ff1f771e4c17117108b172c2ce1952e98738787bb7101f061bf51aa8f03c"}}diff --git a/web-pages/product-site/legacy.py b/web-pages/product-site/legacy.pyindex 5e5a822c1..8510b8553 100644--- a/web-pages/product-site/legacy.py+++ b/web-pages/product-site/legacy.py@@ -54,7 +54,7 @@ def _navigation_markup(language: str, peer_route: str) -> str:f'<div class="nav-links">{links}</div>'f'<a href="{escape(peer_route, quote=True)}" data-language-peer 'f'style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">{peer_label}</a>'- '<a href="https://github.com/modelscope/FunASR" target="_blank" '+ '<a href="/go/github" target="_blank" ''rel="noopener" class="nav-btn">GitHub</a></div></nav>')diff --git a/web-pages/product-site/legacy/blog/funasr-cli-transcribe-command-line.html b/web-pages/product-site/legacy/blog/funasr-cli-transcribe-command-line.htmlindex 70b2fb0b7..f1df2dac6 100644--- a/web-pages/product-site/legacy/blog/funasr-cli-transcribe-command-line.html+++ b/web-pages/product-site/legacy/blog/funasr-cli-transcribe-command-line.html@@ -43,7 +43,7 @@ p code,li code{background:var(--surface);padding:2px 6px;border-radius:5px;font-.post-card .date{color:var(--text-muted);font-size:0.8rem}footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-size:0.8rem}footer a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/funasr-cli-transcribe-command-line.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/funasr-cli-transcribe-command-line.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>用 FunASR 命令行转写音频:一行命令出文本 / JSON / SRT 字幕</h1><p class="post-meta">FunASR 技术博客 · 2026-06-18 · 教程</p>@@ -53,7 +53,7 @@ footer a{color:#94a3b8}</style></head><body>pip install -U funasr # 推荐 funasr ≥ 1.3.26</code></pre><h2>最简单:转成文字</h2><pre><code>funasr audio.wav</code></pre>-<p>直接把识别文本打印到终端。默认用 SenseVoice(非自回归,极快,50+ 语言)。</p>+<p>直接把识别文本打印到终端。默认用 SenseVoice(非自回归,极快,SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种)。</p><h2>生成 SRT 字幕</h2><pre><code>funasr audio.wav -f srt -o ./subs</code></pre><p>在 <code>./subs/</code> 下生成 <code>audio.srt</code>。加 <code>--spk</code> 还能按说话人分句、带真实时间戳:</p>@@ -95,9 +95,9 @@ print(r.text)</code></pre><ul><li><strong>本地、免费、隐私</strong>:无需联网、无 API key。</li><li><strong>快</strong>:SenseVoice 非自回归,远快于 Whisper(<a href="/blog/funasr-vs-whisper-benchmark.html">实测对比</a>),CPU 也能实时。</li>-<li><strong>中文更准</strong> + 50+ 语言;字幕、说话人、热词、JSON 开箱即用。</li>+<li><strong>中文更准</strong> + SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种;字幕、说话人、热词、JSON 开箱即用。</li></ul>-<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">在 GitHub 上 Star FunASR ★</a></div>+<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="/go/github" target="_blank">在 GitHub 上 Star FunASR ★</a></div><!--related-posts--><h2>相关文章</h2><ul class="post-list">@@ -125,5 +125,5 @@ print(r.text)</code></pre><li><a href="/blog/speech-to-text-timestamps-python.html">语音识别带时间戳(字级)</a></li></ul></div></article>-<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="/go/github">GitHub</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/blog/generate-subtitles-srt-vtt-from-audio-video.html b/web-pages/product-site/legacy/blog/generate-subtitles-srt-vtt-from-audio-video.htmlindex 7811dad2a..0334b8d90 100644--- a/web-pages/product-site/legacy/blog/generate-subtitles-srt-vtt-from-audio-video.html+++ b/web-pages/product-site/legacy/blog/generate-subtitles-srt-vtt-from-audio-video.html@@ -1,12 +1,12 @@<!DOCTYPE html><html lang="zh"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">-<title>用 FunASR 自动生成字幕:从音频/视频一键导出 SRT 和 VTT | FunASR</title><meta name="description" content="本地、免费给视频/音频生成字幕:`funasr video.wav -f srt` 一行出 SRT,Python 还能同时导出 VTT、带说话人和真实时间戳。中文强、50+ 语言,无需上传云端。代码全部实测。"><meta name="keywords" content="自动生成字幕, 视频转字幕, 音频生成srt, 字幕生成器, srt 生成 python, vtt 字幕, 视频加字幕, 语音转字幕, 给视频配字幕, whisper 字幕替代">-<link rel="icon" href="/logo.png"><meta property="og:title" content="用 FunASR 自动生成字幕:从音频/视频一键导出 SRT 和 VTT"><meta property="og:description" content="本地、免费给视频/音频生成字幕:`funasr video.wav -f srt` 一行出 SRT,Python 还能同时导出 VTT、带说话人和真实时间戳。中文强、50+ 语言,无需上传云端。代码全部实测。">+<title>用 FunASR 自动生成字幕:从音频/视频一键导出 SRT 和 VTT | FunASR</title><meta name="description" content="本地、免费给视频/音频生成字幕:`funasr video.wav -f srt` 一行出 SRT,Python 还能同时导出 VTT、带说话人和真实时间戳。中文强、SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种,无需上传云端。代码全部实测。"><meta name="keywords" content="自动生成字幕, 视频转字幕, 音频生成srt, 字幕生成器, srt 生成 python, vtt 字幕, 视频加字幕, 语音转字幕, 给视频配字幕, whisper 字幕替代">+<link rel="icon" href="/logo.png"><meta property="og:title" content="用 FunASR 自动生成字幕:从音频/视频一键导出 SRT 和 VTT"><meta property="og:description" content="本地、免费给视频/音频生成字幕:`funasr video.wav -f srt` 一行出 SRT,Python 还能同时导出 VTT、带说话人和真实时间戳。中文强、SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种,无需上传云端。代码全部实测。"><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html"><meta property="og:image" content="https://www.funasr.com/logo.png"><link rel="canonical" href="https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link rel="alternate" hreflang="en" href="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link rel="alternate" hreflang="zh" href="https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"用 FunASR 自动生成字幕:从音频/视频一键导出 SRT 和 VTT","description":"本地、免费给视频/音频生成字幕:`funasr video.wav -f srt` 一行出 SRT,Python 还能同时导出 VTT、带说话人和真实时间戳。中文强、50+ 语言,无需上传云端。代码全部实测。","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"用 FunASR 自动生成字幕:从音频/视频一键导出 SRT 和 VTT","description":"本地、免费给视频/音频生成字幕:`funasr video.wav -f srt` 一行出 SRT,Python 还能同时导出 VTT、带说话人和真实时间戳。中文强、SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种,无需上传云端。代码全部实测。","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -43,11 +43,11 @@ p code,li code{background:var(--surface);padding:2px 6px;border-radius:5px;font-.post-card .date{color:var(--text-muted);font-size:0.8rem}footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-size:0.8rem}footer a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>用 FunASR 自动生成字幕:从音频/视频一键导出 SRT 和 VTT</h1><p class="post-meta">FunASR 技术博客 · 2026-06-18 · 教程</p>-<p>要给视频或音频配字幕,不一定要上传到云端或买 API。FunASR 可以在本地、免费地把语音转成带时间戳的 <strong>SRT</strong> 或 <strong>VTT</strong> 字幕,还能标注说话人。中文尤其强,同时支持 50+ 语言。下面的命令和代码都已实测。</p>+<p>要给视频或音频配字幕,不一定要上传到云端或买 API。FunASR 可以在本地、免费地把语音转成带时间戳的 <strong>SRT</strong> 或 <strong>VTT</strong> 字幕,还能标注说话人。中文尤其强,同时支持 SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种。下面的命令和代码都已实测。</p><h2>第一步(视频):先抽出音频</h2><p>如果源文件是视频,用 ffmpeg 抽出 16kHz 单声道音频(FunASR 的标准输入):</p>@@ -100,9 +100,9 @@ Speaker 0: 欢迎大家来体验达摩院推出的语音识别模型</code></pre<ul><li><strong>本地、免费、隐私</strong>:无需上传、无 API key、无时长费用。</li><li><strong>快</strong>:SenseVoice 非自回归,远快于 Whisper(<a href="/blog/funasr-vs-whisper-benchmark.html">实测对比</a>),长视频也能很快出字幕。</li>-<li><strong>中文更准</strong> + 50+ 语言;自带 VAD 分句、说话人标注、真实时间戳。</li>+<li><strong>中文更准</strong> + SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种;自带 VAD 分句、说话人标注、真实时间戳。</li></ul>-<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">在 GitHub 上 Star FunASR ★</a></div>+<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="/go/github" target="_blank">在 GitHub 上 Star FunASR ★</a></div><!--related-posts--><h2>相关文章</h2><ul class="post-list">@@ -130,5 +130,5 @@ Speaker 0: 欢迎大家来体验达摩院推出的语音识别模型</code></pre<li><a href="/blog/speech-to-text-timestamps-python.html">语音识别带时间戳(字级)</a></li></ul></div></article>-<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="/go/github">GitHub</a></div></footer></body></html>\ No newline at end of filediff --git a/web-pages/product-site/legacy/blog/index.html b/web-pages/product-site/legacy/blog/index.htmlindex 2555d3f4f..07f356bc5 100644--- a/web-pages/product-site/legacy/blog/index.html+++ b/web-pages/product-site/legacy/blog/index.html@@ -47,8 +47,8 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}@media(max-width:900px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}.launch-feature a{grid-template-columns:1fr}.launch-feature .action{white-space:normal}.previous-release .container{grid-template-columns:1fr}}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><section class="launch-feature"><a href="/blog/funasr-v1-4-0-pypi-release.html"><div><span class="date">2026-07-31 · 正式 Release</span><h2>FunASR v1.4.0:更完整的 PyPI 安装包与更安全的 AutoModel 参数</h2><p>修复 SenseVoice 与 RWKV-BAT 包数据,提前拒绝 vda_model 误拼写,并提供 12 个带 SHA-256 的发布资产。</p></div><span class="action">查看发布与校验指南 →</span></a></section><section class="previous-release"><div class="container"><a class="post-card" href="/blog/subtitle-edit-fun-asr-sensevoice-local-subtitles.html"><span class="date">2026-08-03 · 生态集成</span><h2>Subtitle Edit 5.2:Fun-ASR Nano / SenseVoice 本地视频字幕</h2><p>在 Windows、macOS 与 Linux 的成熟字幕工作区中直接选择本地 ASR 引擎,支持 Q4、Q8、F16。</p></a><a class="post-card" href="/blog/funclip-v2-1-0-video-clipping-release.html"><span class="date">2026-07-24</span><h2>FunClip v2.1.0:本地 AI 视频剪辑的首个版本化发布</h2><p>Paraformer、Fun-ASR-Nano、SenseVoice 与大模型剪辑,提供可校验、可回退的源码下载。</p></a></div></section>-<article><div class="container"><h1>FunASR 技术博客</h1><p class="post-meta">语音识别部署、教程与最佳实践</p><a class="post-card" href="/blog/speech-to-text-timestamps-python.html"><span class="date">2026-06-23</span><h2>语音识别带时间戳(字级 timestamps)Python 实战:每个字精确到毫秒</h2><p>FunASR Paraformer 原生字级时间戳:每个字带 [起始毫秒,结束毫秒],一次调用即有。可做逐字高亮、点击跳转、字幕对齐。附真实实测+配对代码。</p></a><a class="post-card" href="/blog/punctuation-restoration-python.html"><span class="date">2026-06-23</span><h2>标点恢复 Python 实战:给无标点文本/ASR 结果自动加标点</h2><p>FunASR ct-punc 开源标点恢复,中英双语,3 行 Python 补全 ,。?(英文还做句首大写),也能一行挂到 ASR 上。附真实实测。</p></a><a class="post-card" href="/blog/chinese-speech-recognition.html"><span class="date">2026-06-22</span><h2>中文语音识别(普通话)Python 实战:用 FunASR 又快又准</h2><p>FunASR 专为中文打造:默认旗舰 Fun-ASR-Nano(CER 8.06%),CPU 用 SenseVoice(7.81%)/ Paraformer(10.18%,时间戳/热词),都远好于 Whisper(~20%)。3 行 Python。</p></a><a class="post-card" href="/blog/self-hosted-google-aws-azure-speech-to-text.html"><span class="date">2026-06-22</span><h2>自托管语音转文字:Google / AWS / Azure 云语音 API 的免费开源替代</h2><p>开源免费(MIT)、本地推理、不按分钟计费、数据不出内网、中文强,OpenAI 兼容改 base_url 即可迁移。附实测代码 + FunASR vs 云 API 对比 + 成本分析。</p></a><a class="post-card" href="/blog/voice-activity-detection-python.html"><span class="date">2026-06-22</span><h2>Python 语音活动检测(VAD):检测语音、去静音、按停顿切分音频</h2><p>FunASR fsmn-vad 3 行代码返回每段语音的起止毫秒。实测 13s 录音 0.12s 切成 2 段、去除 18% 静音;可去静音、切分长录音、给 Whisper 等做预处理防幻觉。</p></a><a class="post-card" href="/blog/japanese-speech-recognition.html"><span class="date">2026-06-22</span><h2>日语语音识别:SenseVoice 一个模型搞定日语转写+标点+情感</h2><p>同一段日语音频实测:SenseVoice 写对同音字 転売、自动加标点,Whisper-small 误为 天売、无标点。原生 ja 支持 + 自动语种识别 + 情感事件,3 行 Python。</p></a><a class="post-card" href="/blog/self-hosted-deepgram-assemblyai-alternative.html"><span class="date">2026-06-23</span><h2>自托管替代 Deepgram / AssemblyAI</h2><p>开源免费、自托管、音频不出本地、中文领先,自带 OpenAI 兼容 API——客户端只改 base_url 即可替代按分钟付费的云 STT。</p></a><a class="post-card" href="/blog/which-funasr-model.html"><span class="date">2026-06-23</span><h2>选哪个 FunASR 模型?Nano vs MLT-Nano vs SenseVoice vs Paraformer</h2><p>模型选型表 + 场景代码:中英日及中文方言/口音选旗舰 Nano,31 语种选独立 MLT-Nano,CPU 再选 SenseVoice 或 Paraformer。</p></a><a class="post-card" href="/blog/lightweight-speech-recognition-cpu.html"><span class="date">2026-06-22</span><h2>轻量语音识别:CPU 上约 250MB 跑中文 ASR</h2><p>单二进制 + 254MB q8 模型,无需 GPU/Python,CPU 0.16s,中文 CER 7.99%——比 whisper.cpp small 还小且准 3 倍。</p></a><a class="post-card" href="/blog/cantonese-speech-recognition.html"><span class="date">2026-06-21</span><h2>粤语语音识别:SenseVoice 原生支持粤语口语(Whisper 会转成普通话)</h2><p>同一段粤语音频实测:SenseVoice 保留 呢/唔/嘅,Whisper 转成普通话书面语。原生 yue 支持 + 自动语种识别,3 行 Python。</p></a><a class="post-card" href="/blog/funasr-vs-faster-whisper-chinese.html"><span class="date">2026-06-21</span><h2>FunASR vs faster-whisper:中文与粤语实测对比</h2><p>粤语被 faster-whisper 误判为普通话、日语同音字错;SenseVoice 原生支持粤语+语种识别,中文 CER 低约 2.7 倍。实测。</p></a><a class="post-card" href="/blog/funasr-llama-cpp-whisper-cpp-alternative.html"><span class="date">2026-06-20</span><h2>FunASR 跑进 llama.cpp:中文 ASR 的 whisper.cpp 替代品(CPU/零 Python)</h2><p>单个自包含二进制、内置 VAD、吃任意音频,下载即用转写中文;中文 CPU 上比 whisper.cpp 准约 2.7 倍。3 步实测。</p></a><a class="post-card" href="/blog/speech-to-text-python-transcribe-audio.html"><span class="date">2026-06-18</span><h2>Python 语音转文字:用 FunASR 本地免费转写音频</h2><p>几行 Python 把音频转成文本,带时间戳/说话人/批量;本地、免费、无 API key、中文强。</p></a><a class="post-card" href="/blog/generate-subtitles-srt-vtt-from-audio-video.html"><span class="date">2026-06-18</span><h2>用 FunASR 自动生成字幕:音频/视频一键出 SRT 和 VTT</h2><p>一行命令出 SRT,Python 同时导出 VTT,带说话人和真实时间戳;本地、免费、中文强。</p></a><a class="post-card" href="/blog/self-hosted-openai-whisper-api-alternative.html"><span class="date">2026-06-18</span><h2>自托管 OpenAI Whisper API 替代:FunASR 起兼容 /v1/audio/transcriptions 服务</h2><p>funasr-server 暴露 OpenAI 兼容接口,OpenAI SDK 只改 base_url 就能用;本地、免费、隐私、中文更准。</p></a><a class="post-card" href="/blog/funasr-cli-transcribe-command-line.html"><span class="date">2026-06-18</span><h2>用 FunASR 命令行转写音频:文本/JSON/SRT 字幕</h2><p>一行命令出文字/字幕/JSON,--spk 带说话人;还能 funasr-server 起 OpenAI 兼容 API。</p></a><a class="post-card" href="/blog/funasr-transcribe-long-audio.html"><span class="date">2026-06-17</span><h2>用 FunASR 转写超长音频:1 小时一次搞定</h2><p>Whisper 限 30 秒,FunASR 内置 VAD 一次吃下任意时长;实测 13 分钟 4.3 秒转完(186x)。</p></a><a class="post-card" href="/blog/funasr-realtime-streaming-asr.html"><span class="date">2026-06-17</span><h2>用 FunASR 实现实时流式语音识别(边说边出字)</h2><p>600ms 级低延迟流式 ASR:分块+cache 边说边出字,含 2-pass(流式+离线)最佳实践。</p></a><a class="post-card" href="/blog/sensevoice-emotion-language-detection.html"><span class="date">2026-06-17</span><h2>超越转写:用 SenseVoice 识别语言、情感与声学事件</h2><p>一次非自回归前向同时输出转写+语种+情感+音频事件,Whisper 做不到的四合一。</p></a><a class="post-card" href="/blog/funasr-speaker-diarization.html"><span class="date">2026-06-17</span><h2>用 FunASR 做说话人分离:谁在何时说了什么</h2><p>一次 generate 调用同时输出转写+说话人标签+时间戳,替代 pyannote+Whisper,无需 HF 授权。</p></a><a class="post-card" href="/blog/funasr-vs-whisper-benchmark.html"><span class="date">2026-06-16</span><h2>FunASR vs Whisper 实测对比:谁更快更准</h2><p>184 中文文件 H100 实测:SenseVoice 169.6x、CER 7.81%,完整速度+准确率数据。</p></a><a class="post-card" href="/blog/fun-asr-nano-guide.html"><span class="date">2026-06-16</span><h2>Fun-ASR-Nano 使用指南:800M 端到端语音识别大模型</h2><p>主力旗舰,中英日 + 7 大中文方言/26 种口音,热词/流式/说话人分离;31 语种请用 MLT-Nano。</p></a><a class="post-card" href="/blog/sensevoice-deployment-guide.html"><span class="date">2026-06-16</span><h2>SenseVoice 部署指南:比 Whisper 快 15 倍</h2><p>3 行代码跑通多语言识别,含语种/情感/事件检测、VAD、GPU/CPU。</p></a><p>更多:<a href="/quickstart.html">快速上手</a> · <a href="/models.html">模型</a></p></div></article>+<article><div class="container"><h1>FunASR 技术博客</h1><p class="post-meta">语音识别部署、教程与最佳实践</p><a class="post-card" href="/blog/speech-to-text-timestamps-python.html"><span class="date">2026-06-23</span><h2>语音识别带时间戳(字级 timestamps)Python 实战:每个字精确到毫秒</h2><p>FunASR Paraformer 原生字级时间戳:每个字带 [起始毫秒,结束毫秒],一次调用即有。可做逐字高亮、点击跳转、字幕对齐。附真实实测+配对代码。</p></a><a class="post-card" href="/blog/punctuation-restoration-python.html"><span class="date">2026-06-23</span><h2>标点恢复 Python 实战:给无标点文本/ASR 结果自动加标点</h2><p>FunASR ct-punc 开源标点恢复,中英双语,3 行 Python 补全 ,。?(英文还做句首大写),也能一行挂到 ASR 上。附真实实测。</p></a><a class="post-card" href="/blog/chinese-speech-recognition.html"><span class="date">2026-06-22</span><h2>中文语音识别(普通话)Python 实战:用 FunASR 又快又准</h2><p>FunASR 专为中文打造:默认旗舰 Fun-ASR-Nano(CER 8.06%),CPU 用 SenseVoice(7.81%)/ Paraformer(10.18%,时间戳/热词),都远好于 Whisper(~20%)。3 行 Python。</p></a><a class="post-card" href="/blog/self-hosted-google-aws-azure-speech-to-text.html"><span class="date">2026-06-22</span><h2>自托管语音转文字:Google / AWS / Azure 云语音 API 的免费开源替代</h2><p>开源免费(MIT)、本地推理、不按分钟计费、数据不出内网、中文强,OpenAI 兼容改 base_url 即可迁移。附实测代码 + FunASR vs 云 API 对比 + 成本分析。</p></a><a class="post-card" href="/blog/voice-activity-detection-python.html"><span class="date">2026-06-22</span><h2>Python 语音活动检测(VAD):检测语音、去静音、按停顿切分音频</h2><p>FunASR fsmn-vad 3 行代码返回每段语音的起止毫秒。实测 13s 录音 0.12s 切成 2 段、去除 18% 静音;可去静音、切分长录音、给 Whisper 等做预处理防幻觉。</p></a><a class="post-card" href="/blog/japanese-speech-recognition.html"><span class="date">2026-06-22</span><h2>日语语音识别:SenseVoice 一个模型搞定日语转写+标点+情感</h2><p>同一段日语音频实测:SenseVoice 写对同音字 転売、自动加标点,Whisper-small 误为 天売、无标点。原生 ja 支持 + 自动语种识别 + 情感事件,3 行 Python。</p></a><a class="post-card" href="/blog/self-hosted-deepgram-assemblyai-alternative.html"><span class="date">2026-06-23</span><h2>自托管替代 Deepgram / AssemblyAI</h2><p>开源免费、自托管、音频不出本地、中文领先,自带 OpenAI 兼容 API——客户端只改 base_url 即可替代按分钟付费的云 STT。</p></a><a class="post-card" href="/blog/which-funasr-model.html"><span class="date">2026-06-23</span><h2>选哪个 FunASR 模型?Nano vs MLT-Nano vs SenseVoice vs Paraformer</h2><p>模型选型表 + 场景代码:中英日及中文方言/口音选旗舰 Nano,31 语种选独立 MLT-Nano,CPU 再选 SenseVoice 或 Paraformer。</p></a><a class="post-card" href="/blog/lightweight-speech-recognition-cpu.html"><span class="date">2026-06-22</span><h2>轻量语音识别:CPU 上约 250MB 跑中文 ASR</h2><p>单二进制 + 254MB q8 模型,无需 GPU/Python,CPU 0.16s,中文 CER 7.99%——比 whisper.cpp small 还小且准 3 倍。</p></a><a class="post-card" href="/blog/cantonese-speech-recognition.html"><span class="date">2026-06-21</span><h2>粤语语音识别:SenseVoice 原生支持粤语口语(Whisper 会转成普通话)</h2><p>同一段粤语音频实测:SenseVoice 保留 呢/唔/嘅,Whisper 转成普通话书面语。原生 yue 支持 + 自动语种识别,3 行 Python。</p></a><a class="post-card" href="/blog/funasr-vs-faster-whisper-chinese.html"><span class="date">2026-06-21</span><h2>FunASR vs faster-whisper:中文与粤语实测对比</h2><p>粤语被 faster-whisper 误判为普通话、日语同音字错;SenseVoice 原生支持粤语+语种识别,中文 CER 低约 2.7 倍。实测。</p></a><a class="post-card" href="/blog/funasr-llama-cpp-whisper-cpp-alternative.html"><span class="date">2026-06-20</span><h2>FunASR 跑进 llama.cpp:中文 ASR 的 whisper.cpp 替代品(CPU/零 Python)</h2><p>单个自包含二进制、内置 VAD、吃任意音频,下载即用转写中文;中文 CPU 上比 whisper.cpp 准约 2.7 倍。3 步实测。</p></a><a class="post-card" href="/blog/speech-to-text-python-transcribe-audio.html"><span class="date">2026-06-18</span><h2>Python 语音转文字:用 FunASR 本地免费转写音频</h2><p>几行 Python 把音频转成文本,带时间戳/说话人/批量;本地、免费、无 API key、中文强。</p></a><a class="post-card" href="/blog/generate-subtitles-srt-vtt-from-audio-video.html"><span class="date">2026-06-18</span><h2>用 FunASR 自动生成字幕:音频/视频一键出 SRT 和 VTT</h2><p>一行命令出 SRT,Python 同时导出 VTT,带说话人和真实时间戳;本地、免费、中文强。</p></a><a class="post-card" href="/blog/self-hosted-openai-whisper-api-alternative.html"><span class="date">2026-06-18</span><h2>自托管 OpenAI Whisper API 替代:FunASR 起兼容 /v1/audio/transcriptions 服务</h2><p>funasr-server 暴露 OpenAI 兼容接口,OpenAI SDK 只改 base_url 就能用;本地、免费、隐私、中文更准。</p></a><a class="post-card" href="/blog/funasr-cli-transcribe-command-line.html"><span class="date">2026-06-18</span><h2>用 FunASR 命令行转写音频:文本/JSON/SRT 字幕</h2><p>一行命令出文字/字幕/JSON,--spk 带说话人;还能 funasr-server 起 OpenAI 兼容 API。</p></a><a class="post-card" href="/blog/funasr-transcribe-long-audio.html"><span class="date">2026-06-17</span><h2>用 FunASR 转写超长音频:1 小时一次搞定</h2><p>Whisper 限 30 秒,FunASR 内置 VAD 一次吃下任意时长;实测 13 分钟 4.3 秒转完(186x)。</p></a><a class="post-card" href="/blog/funasr-realtime-streaming-asr.html"><span class="date">2026-06-17</span><h2>用 FunASR 实现实时流式语音识别(边说边出字)</h2><p>600ms 级低延迟流式 ASR:分块+cache 边说边出字,含 2-pass(流式+离线)最佳实践。</p></a><a class="post-card" href="/blog/sensevoice-emotion-language-detection.html"><span class="date">2026-06-17</span><h2>超越转写:用 SenseVoice 识别语言、情感与声学事件</h2><p>一次非自回归前向同时输出转写+语种+情感+音频事件,Whisper 做不到的四合一。</p></a><a class="post-card" href="/blog/funasr-speaker-diarization.html"><span class="date">2026-06-17</span><h2>用 FunASR 做说话人分离:谁在何时说了什么</h2><p>一次 generate 调用同时输出转写+说话人标签+时间戳,替代 pyannote+Whisper,无需 HF 授权。</p></a><a class="post-card" href="/blog/funasr-vs-whisper-benchmark.html"><span class="date">2026-06-16</span><h2>FunASR vs Whisper 实测对比:谁更快更准</h2><p>184 中文文件 H100 实测:SenseVoice 169.6x、CER 7.81%,完整速度+准确率数据。</p></a><a class="post-card" href="/blog/fun-asr-nano-guide.html"><span class="date">2026-06-16</span><h2>Fun-ASR-Nano 使用指南:800M 端到端语音识别大模型</h2><p>主力旗舰,中英日 + 7 大中文方言/26 种口音,热词/流式/说话人分离;31 语种请用 MLT-Nano。</p></a><a class="post-card" href="/blog/sensevoice-deployment-guide.html"><span class="date">2026-06-16</span><h2>SenseVoice 部署指南:五语种识别、情感与音频事件</h2><p>3 行代码跑通多语言识别,含语种/情感/事件检测、VAD、GPU/CPU。</p></a><p>更多:<a href="/quickstart.html">快速上手</a> · <a href="/models.html">模型</a></p></div></article><footer><div class="container">© 2024-2026 Tongyi Lab, Alibaba Group · <a href="https://github.com/modelscope/FunASR/blob/main/LICENSE">MIT License</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/blog/japanese-speech-recognition.html b/web-pages/product-site/legacy/blog/japanese-speech-recognition.htmlindex c6a8b2c98..308f2cf41 100644--- a/web-pages/product-site/legacy/blog/japanese-speech-recognition.html+++ b/web-pages/product-site/legacy/blog/japanese-speech-recognition.html@@ -57,7 +57,7 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/japanese-speech-recognition.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/japanese-speech-recognition.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>日语语音识别:SenseVoice 一个模型搞定日语转写 + 标点 + 情感(附与 Whisper 的真实对比)</h1><p class="post-meta">2026-06-22 · FunASR 团队</p>@@ -106,12 +106,12 @@ print(rich_transcription_postprocess(res[0]["text"]))<tr><td>情感 / 音频事件</td><td>✅ 一次出</td><td>❌</td></tr><tr><td>逆文本归一化(ITN)</td><td>✅ 内置</td><td>部分</td></tr><tr><td>速度</td><td>非自回归,实测 RTF≈0.04(GPU)</td><td>自回归,更慢</td></tr>-<tr><td>多语种</td><td>一个模型覆盖 50+ 语种(ja/ko/zh/yue/en…)</td><td>多语种</td></tr>+<tr><td>多语种</td><td>一个模型覆盖 中、粤、英、日、韩五种语言(ja/ko/zh/yue/en…)</td><td>多语种</td></tr><tr><td>许可证</td><td>开源,可商用</td><td>开源</td></tr></table><p>如果你的场景是<strong>日语 + 中文 + 英文混合</strong>(跨境电商客服、字幕、会议),SenseVoice 一个模型就能全覆盖,还自带情感分析。想了解它的情感/事件检测能力,看 <a href="/blog/sensevoice-emotion-language-detection.html">SenseVoice 情感与音频事件识别</a>;不确定选哪个模型,看 <a href="/blog/which-funasr-model.html">FunASR 模型选型指南</a>;中文准确率对比见 <a href="/blog/funasr-vs-whisper-benchmark.html">FunASR vs Whisper benchmark</a>。</p>-<div class="cta"><p><strong>FunASR 全家桶都是开源的</strong>——工业级 ASR / VAD / 标点 / 说话人 / 情感事件 / LLM-ASR,日语等 50+ 语种开箱即用。觉得有用就点个 Star 支持一下 👇</p><p><a class="btn" href="https://github.com/QwenAudio/SenseVoice" target="_blank">⭐ Star SenseVoice</a></p><p style="margin-top:10px;font-size:0.85rem">也欢迎 Star:<a href="https://github.com/modelscope/FunASR" target="_blank">FunASR</a> · <a href="https://github.com/QwenAudio/Fun-ASR" target="_blank">Fun-ASR</a> · <a href="https://github.com/modelscope/FunClip" target="_blank">FunClip</a></p></div>+<div class="cta"><p><strong>FunASR 全家桶都是开源的</strong>——工业级 ASR / VAD / 标点 / 说话人 / 情感事件 / LLM-ASR,日语等 中、粤、英、日、韩五种语言开箱即用。觉得有用就点个 Star 支持一下 👇</p><p><a class="btn" href="/go/sensevoice" target="_blank">⭐ Star SenseVoice</a></p><p style="margin-top:10px;font-size:0.85rem">也欢迎 Star:<a href="/go/github" target="_blank">FunASR</a> · <a href="/go/fun-asr" target="_blank">Fun-ASR</a> · <a href="/go/funclip" target="_blank">FunClip</a></p></div><!--related-posts--><h2>相关文章</h2><ul class="post-list">diff --git a/web-pages/product-site/legacy/blog/self-hosted-openai-whisper-api-alternative.html b/web-pages/product-site/legacy/blog/self-hosted-openai-whisper-api-alternative.htmlindex e639a72a0..46eafe660 100644--- a/web-pages/product-site/legacy/blog/self-hosted-openai-whisper-api-alternative.html+++ b/web-pages/product-site/legacy/blog/self-hosted-openai-whisper-api-alternative.html@@ -45,7 +45,7 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}@media(max-width:900px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.75rem}table{table-layout:fixed;font-size:0.78rem}th,td{padding:8px 4px;overflow-wrap:anywhere}p code,li code{overflow-wrap:anywhere;word-break:break-word}}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/self-hosted-openai-whisper-api-alternative.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/self-hosted-openai-whisper-api-alternative.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>自托管 OpenAI Whisper API 替代方案:FunASR 一行起兼容 /v1/audio/transcriptions 服务</h1><p class="post-meta">FunASR 技术博客 · 2026-06-18 · 教程</p>@@ -89,7 +89,7 @@ print(r.text)<h2>选哪个模型</h2><ul>-<li><code>sensevoice</code> —— 默认,非自回归,极快,50+ 语言,带情感/事件;日常首选。</li>+<li><code>sensevoice</code> —— 默认,非自回归,极快,中、粤、英、日、韩五种语言,带情感/事件;日常首选。</li><li><code>fun-asr-nano</code> —— LLM 解码,面向中英日及中文方言/口音的旗舰模型(需要 vLLM)。</li><li>需要 31 语种 —— 部署独立的 <code>FunAudioLLM/Fun-ASR-MLT-Nano-2512</code> checkpoint。</li><li><code>paraformer</code> —— 经典中文生产级,稳。</li>@@ -119,7 +119,7 @@ AUDIO_STT_OPENAI_API_BASE_URL=http://host.docker.internal:8000/v1AUDIO_STT_OPENAI_API_KEY=funasrAUDIO_STT_MODEL=sensevoice</pre><p>之后 Open-WebUI 的麦克风语音输入就走 FunASR 了——中文转写更准、数据不出本地。同样的配置适用于任何"支持自定义 OpenAI STT 端点"的工具(LibreChat、AnythingLLM 等)。</p>-<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">在 GitHub 上 Star FunASR ★</a></div>+<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="/go/github" target="_blank">在 GitHub 上 Star FunASR ★</a></div><!--related-posts--><h2>相关文章</h2><ul class="post-list">@@ -147,5 +147,5 @@ AUDIO_STT_MODEL=sensevoice</pre><li><a href="/blog/speech-to-text-timestamps-python.html">语音识别带时间戳(字级)</a></li></ul></div></article>-<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="/go/github">GitHub</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/blog/sensevoice-deployment-guide.html b/web-pages/product-site/legacy/blog/sensevoice-deployment-guide.htmlindex 0bfb1303a..909dc07fd 100644--- a/web-pages/product-site/legacy/blog/sensevoice-deployment-guide.html+++ b/web-pages/product-site/legacy/blog/sensevoice-deployment-guide.html@@ -3,12 +3,12 @@<head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">-<title>SenseVoice 部署指南:比 Whisper 快 15 倍的多语言语音识别 | FunASR 博客</title>-<meta name="description" content="SenseVoice 部署完整教程:3 行 Python 代码跑通多语言语音识别,含语种/情感/音频事件检测、VAD 长音频分段、GPU/CPU 选择。比 Whisper 快 15 倍的开源 ASR。">+<title>SenseVoice 部署指南:五语种识别、情感与音频事件 | FunASR 博客</title>+<meta name="description" content="SenseVoice 部署完整教程:3 行 Python 代码跑通中、粤、英、日、韩语音识别,含语种、情感、音频事件检测、VAD 长音频分段及 GPU/CPU 选择。"><meta name="keywords" content="SenseVoice, SenseVoice 部署, 语音识别, 多语言语音识别, Whisper 替代, FunASR, 情感识别, 语种识别, ASR 教程, 开源语音识别"><link rel="icon" href="/logo.png">-<meta property="og:title" content="SenseVoice 部署指南:比 Whisper 快 15 倍的多语言语音识别 | FunASR 博客">-<meta property="og:description" content="SenseVoice 部署完整教程:3 行 Python 代码跑通多语言语音识别,含语种/情感/音频事件检测、VAD 长音频分段、GPU/CPU 选择。比 Whisper 快 15 倍的开源 ASR。">+<meta property="og:title" content="SenseVoice 部署指南:五语种识别、情感与音频事件 | FunASR 博客">+<meta property="og:description" content="SenseVoice 部署完整教程:3 行 Python 代码跑通中、粤、英、日、韩语音识别,含语种、情感、音频事件检测、VAD 长音频分段及 GPU/CPU 选择。"><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/blog/sensevoice-deployment-guide.html"><meta property="og:image" content="https://www.funasr.com/logo.png">@@ -18,7 +18,7 @@<link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/sensevoice-deployment-guide.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"SenseVoice 部署指南:比 Whisper 快 15 倍的多语言语音识别 | FunASR 博客","description":"SenseVoice 部署完整教程:3 行 Python 代码跑通多语言语音识别,含语种/情感/音频事件检测、VAD 长音频分段、GPU/CPU 选择。比 Whisper 快 15 倍的开源 ASR。","datePublished":"2026-06-16","dateModified":"2026-06-16","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/sensevoice-deployment-guide.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"SenseVoice 部署指南:五语种识别、情感与音频事件 | FunASR 博客","description":"SenseVoice 部署完整教程:3 行 Python 代码跑通中、粤、英、日、韩语音识别,含语种、情感、音频事件检测、VAD 长音频分段及 GPU/CPU 选择。","datePublished":"2026-06-16","dateModified":"2026-08-17","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/sensevoice-deployment-guide.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -54,14 +54,18 @@ p code,li code{background:var(--surface);padding:2px 6px;border-radius:5px;font-.post-card p{margin:0;color:var(--text-soft);font-size:0.9rem}.post-card .date{color:var(--text-muted);font-size:0.8rem}footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-size:0.8rem}-footer a{color:#94a3b8}</style>+footer a{color:#94a3b8}+@media(max-width:700px){+.nav{padding:12px 0}.nav .container{gap:16px}.nav-links{display:none}.nav-btn{white-space:nowrap}+article{padding-top:96px}+}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/sensevoice-deployment-guide.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/sensevoice-deployment-guide.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container">-<h1>SenseVoice 部署指南:比 Whisper 快 15 倍的多语言语音识别</h1>+<h1>SenseVoice 部署指南:五语种识别、情感与音频事件</h1><p class="post-meta">FunASR 技术博客 · 2026-06-16 · 部署教程</p>-<p><strong>SenseVoice</strong> 是 FunAudioLLM 团队开源的多语言语音理解大模型。它不仅做语音识别(ASR),还能在一次前向推理中同时输出<strong>语种识别、情感识别、音频事件检测</strong>。模型采用非自回归结构,推理速度比 Whisper-Large 快约 <strong>15 倍</strong>,单条 10 秒音频的识别延迟可低至几十毫秒,非常适合工业级实时与批量场景。</p>+<p><strong>SenseVoice</strong> 是 FunAudioLLM 团队开源的语音理解模型,支持普通话、粤语、英语、日语和韩语。它不仅做语音识别(ASR),还能在一次前向推理中同时输出<strong>语种识别、情感识别、音频事件检测</strong>。模型采用非自回归结构,适合低延迟和批量场景;实际吞吐和延迟取决于硬件、音频长度及 batch,需要在目标环境实测。</p><p>本文用可直接复制运行的代码,带你 3 步把 SenseVoice 跑起来。</p><h2>1. 安装</h2>@@ -112,8 +116,8 @@ res = model.generate(input="long_audio.wav", language="auto", use_itn=True, batc<table><thead><tr><th>维度</th><th>SenseVoice-Small</th></tr></thead><tbody>-<tr><td>相对 Whisper-Large 速度</td><td>约 15×</td></tr>-<tr><td>支持语言</td><td>中文 / 粤语 / 英文 / 日文 / 韩文 等</td></tr>+<tr><td>性能验证</td><td>在目标硬件、音频分布和 batch 下实测 RTF 与 p50/p95</td></tr>+<tr><td>支持语言</td><td>中文 / 粤语 / 英文 / 日文 / 韩文</td></tr><tr><td>附加能力</td><td>语种识别 · 情感识别 · 音频事件检测</td></tr><tr><td>架构</td><td>非自回归(低延迟)</td></tr><tr><td>许可证</td><td>开源(可商用)</td></tr>@@ -123,7 +127,7 @@ res = model.generate(input="long_audio.wav", language="auto", use_itn=True, batc<div class="cta"><h3>开始使用 SenseVoice</h3><p>如果这篇指南对你有帮助,欢迎到 GitHub 点个 Star ⭐ 支持开源。</p>-<a href="https://github.com/QwenAudio/SenseVoice" class="btn">SenseVoice GitHub ★</a>+<a href="/go/sensevoice" class="btn">SenseVoice GitHub ★</a></div><p>延伸阅读:<a href="/vs-whisper.html">FunASR vs Whisper 对比</a> · <a href="/quickstart.html">FunASR 快速上手</a> · <a href="/models.html">全部模型</a></p>@@ -156,4 +160,4 @@ res = model.generate(input="long_audio.wav", language="auto", use_itn=True, batc</div></article><footer><div class="container">© 2024-2026 Tongyi Lab, Alibaba Group · <a href="https://github.com/modelscope/FunASR/blob/main/LICENSE">MIT License</a></div></footer></body>-</html>\ No newline at end of file+</html>diff --git a/web-pages/product-site/legacy/blog/sensevoice-emotion-language-detection.html b/web-pages/product-site/legacy/blog/sensevoice-emotion-language-detection.htmlindex 8371733ec..25a7c264f 100644--- a/web-pages/product-site/legacy/blog/sensevoice-emotion-language-detection.html+++ b/web-pages/product-site/legacy/blog/sensevoice-emotion-language-detection.html@@ -4,11 +4,11 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>Python 语音情感识别:一个模型搞定语种 + 情感 + 音频事件(SenseVoice)| FunASR 博客</title>-<meta name="description" content="用 SenseVoice 一次推理同时输出语种识别、情感识别(高兴/愤怒/中性)、音频事件(背景音乐/掌声/笑声)与 ITN。3 行 Python 代码,附 5 语言真实输出,比 Whisper 快 15 倍。">+<meta name="description" content="用 SenseVoice 一次推理同时输出语种识别、情感识别(高兴/愤怒/中性)、音频事件(背景音乐/掌声/笑声)与 ITN。3 行 Python 代码,附 5 语言真实输出,实际吞吐取决于硬件、音频和 batch。"><meta name="keywords" content="语音情感识别, SenseVoice, 语种识别, 音频事件检测, 多语言语音识别, Whisper 替代, FunASR, 情感识别 python, 语音识别, 开源语音识别"><link rel="icon" href="/logo.png"><meta property="og:title" content="Python 语音情感识别:一个模型搞定语种 + 情感 + 音频事件(SenseVoice)| FunASR 博客">-<meta property="og:description" content="用 SenseVoice 一次推理同时输出语种识别、情感识别(高兴/愤怒/中性)、音频事件(背景音乐/掌声/笑声)与 ITN。3 行 Python 代码,附 5 语言真实输出,比 Whisper 快 15 倍。">+<meta property="og:description" content="用 SenseVoice 一次推理同时输出语种识别、情感识别(高兴/愤怒/中性)、音频事件(背景音乐/掌声/笑声)与 ITN。3 行 Python 代码,附 5 语言真实输出,实际吞吐取决于硬件、音频和 batch。"><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/blog/sensevoice-emotion-language-detection.html"><meta property="og:image" content="https://www.funasr.com/logo.png">@@ -18,7 +18,7 @@<link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/sensevoice-emotion-language-detection.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"Python 语音情感识别:一个模型搞定语种 + 情感 + 音频事件(SenseVoice)| FunASR 博客","description":"用 SenseVoice 一次推理同时输出语种识别、情感识别(高兴/愤怒/中性)、音频事件(背景音乐/掌声/笑声)与 ITN。3 行 Python 代码,附 5 语言真实输出,比 Whisper 快 15 倍。","datePublished":"2026-06-19","dateModified":"2026-06-19","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/sensevoice-emotion-language-detection.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"Python 语音情感识别:一个模型搞定语种 + 情感 + 音频事件(SenseVoice)| FunASR 博客","description":"用 SenseVoice 一次推理同时输出语种识别、情感识别(高兴/愤怒/中性)、音频事件(背景音乐/掌声/笑声)与 ITN。3 行 Python 代码,附 5 语言真实输出,实际吞吐取决于硬件、音频和 batch。","datePublished":"2026-06-19","dateModified":"2026-06-19","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/sensevoice-emotion-language-detection.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -57,18 +57,18 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/sensevoice-emotion-language-detection.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/sensevoice-emotion-language-detection.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Python 语音情感识别:一个模型同时输出语种、情感与音频事件(SenseVoice)</h1><p class="post-meta">2026-06-19 · FunASR 团队</p><p>大多数语音识别(ASR)模型只给你<strong>文字</strong>。但真实场景里,你往往还想知道:说话人是<strong>什么情绪</strong>?这段是<strong>哪种语言</strong>?背景是<strong>纯人声还是带音乐</strong>?用 Whisper 你得叠一堆东西——语种检测 + 单独的情感模型 + 音频事件分类器,既慢又难维护。</p>-<p><strong>SenseVoice</strong> 是 FunAudioLLM 团队开源的多语种语音理解模型。它在<strong>一次非自回归前向</strong>里就同时返回:转写文本、<strong>语种识别</strong>、<strong>情感识别</strong>、<strong>音频事件检测</strong>,外加逆文本归一化(ITN)。架构非自回归,<strong>比 Whisper-Large 快约 15 倍</strong>,适合实时也适合大批量。</p>+<p><strong>SenseVoice</strong> 是 FunAudioLLM 团队开源的多语种语音理解模型。它在<strong>一次非自回归前向</strong>里就同时返回:转写文本、<strong>语种识别</strong>、<strong>情感识别</strong>、<strong>音频事件检测</strong>,外加逆文本归一化(ITN)。架构非自回归,适合低延迟和批量推理;实际吞吐取决于硬件、音频长度和 batch。</p><h2>一次推理能拿到什么</h2><table><tr><th>能力</th><th>说明</th></tr>-<tr><td>语音识别(ASR)</td><td>50+ 语种,中文准确率领先</td></tr>+<tr><td>语音识别(ASR)</td><td>中、粤、英、日、韩五种语言,中文准确率领先</td></tr><tr><td>语种识别(LID)</td><td>自动判别 <code>zh / en / ja / ko / yue …</code></td></tr><tr><td>情感识别</td><td>HAPPY 😊 / SAD 😔 / ANGRY 😡 / NEUTRAL / FEARFUL / DISGUSTED / SURPRISED</td></tr><tr><td>音频事件</td><td>Speech / BGM 🎵 / Applause 👏 / Laughter / Cry</td></tr>@@ -125,7 +125,7 @@ print(language, emotion, event, text)</pre></table><p>要的不只是文字,而是对音频的<strong>理解</strong>时,SenseVoice 一个模型就够了——无需再拼语种检测 + 情感模型 + 事件分类器。</p>-<div class="cta"><p><strong>FunASR 全家桶都是开源的</strong>——工业级 ASR / VAD / 标点 / 说话人 / 情感事件 / LLM-ASR。觉得有用就点个 Star 支持一下 👇</p><p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">⭐ Star FunASR</a></p><p style="margin-top:10px;font-size:0.85rem">也欢迎 Star:<a href="https://github.com/QwenAudio/SenseVoice" target="_blank">SenseVoice</a> · <a href="https://github.com/QwenAudio/Fun-ASR" target="_blank">Fun-ASR</a> · <a href="https://github.com/modelscope/FunClip" target="_blank">FunClip</a></p></div>+<div class="cta"><p><strong>FunASR 全家桶都是开源的</strong>——工业级 ASR / VAD / 标点 / 说话人 / 情感事件 / LLM-ASR。觉得有用就点个 Star 支持一下 👇</p><p><a class="btn" href="/go/github" target="_blank">⭐ Star FunASR</a></p><p style="margin-top:10px;font-size:0.85rem">也欢迎 Star:<a href="/go/sensevoice" target="_blank">SenseVoice</a> · <a href="/go/fun-asr" target="_blank">Fun-ASR</a> · <a href="/go/funclip" target="_blank">FunClip</a></p></div><!--related-posts--><h2>相关文章</h2><ul class="post-list">diff --git a/web-pages/product-site/legacy/blog/speech-to-text-python-transcribe-audio.html b/web-pages/product-site/legacy/blog/speech-to-text-python-transcribe-audio.htmlindex 3830c160f..eef47b22a 100644--- a/web-pages/product-site/legacy/blog/speech-to-text-python-transcribe-audio.html+++ b/web-pages/product-site/legacy/blog/speech-to-text-python-transcribe-audio.html@@ -1,12 +1,12 @@<!DOCTYPE html><html lang="zh"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">-<title>Python 语音转文字:用 FunASR 本地、免费地把音频转成文本 | FunASR</title><meta name="description" content="用 Python 几行代码把音频转成文字:`AutoModel(...).generate(...)`,本地运行、免费、无 API key。带时间戳、说话人、批量处理,中文强、50+ 语言。代码全部实测。"><meta name="keywords" content="python 语音转文字, python 语音识别, 音频转文字 python, 语音转文本 python, funasr python, 离线语音识别 python, 免费语音转文字, python 语音转文本教程">-<link rel="icon" href="/logo.png"><meta property="og:title" content="Python 语音转文字:用 FunASR 本地、免费地把音频转成文本"><meta property="og:description" content="用 Python 几行代码把音频转成文字:`AutoModel(...).generate(...)`,本地运行、免费、无 API key。带时间戳、说话人、批量处理,中文强、50+ 语言。代码全部实测。">+<title>Python 语音转文字:用 FunASR 本地、免费地把音频转成文本 | FunASR</title><meta name="description" content="用 Python 几行代码把音频转成文字:`AutoModel(...).generate(...)`,本地运行、免费、无 API key。带时间戳、说话人、批量处理,中文强、SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种。代码全部实测。"><meta name="keywords" content="python 语音转文字, python 语音识别, 音频转文字 python, 语音转文本 python, funasr python, 离线语音识别 python, 免费语音转文字, python 语音转文本教程">+<link rel="icon" href="/logo.png"><meta property="og:title" content="Python 语音转文字:用 FunASR 本地、免费地把音频转成文本"><meta property="og:description" content="用 Python 几行代码把音频转成文字:`AutoModel(...).generate(...)`,本地运行、免费、无 API key。带时间戳、说话人、批量处理,中文强、SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种。代码全部实测。"><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html"><meta property="og:image" content="https://www.funasr.com/logo.png"><link rel="canonical" href="https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html"><link rel="alternate" hreflang="en" href="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"><link rel="alternate" hreflang="zh" href="https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html"><link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"Python 语音转文字:用 FunASR 本地、免费地把音频转成文本","description":"用 Python 几行代码把音频转成文字:`AutoModel(...).generate(...)`,本地运行、免费、无 API key。带时间戳、说话人、批量处理,中文强、50+ 语言。代码全部实测。","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"Python 语音转文字:用 FunASR 本地、免费地把音频转成文本","description":"用 Python 几行代码把音频转成文字:`AutoModel(...).generate(...)`,本地运行、免费、无 API key。带时间戳、说话人、批量处理,中文强、SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种。代码全部实测。","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -45,11 +45,11 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}@media(max-width:900px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.75rem}table{table-layout:fixed;font-size:0.78rem}th,td{padding:8px 4px;overflow-wrap:anywhere}p code,li code{overflow-wrap:anywhere;word-break:break-word}}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="/quickstart.html">快速上手</a><a href="/models.html">模型</a><a href="/ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="/vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Python 语音转文字:用 FunASR 本地、免费地把音频转成文本</h1><p class="post-meta">FunASR 技术博客 · 2026-06-18 · 教程</p>-<p>在 Python 里把音频转成文字,不一定要用云 API 或庞大的 Whisper。FunASR 几行代码就能在本地、免费地转写,还自带时间戳、说话人分离和批量处理。中文尤其强,同时支持 50+ 语言。下面每段代码都已实测。</p>+<p>在 Python 里把音频转成文字,不一定要用云 API 或庞大的 Whisper。FunASR 几行代码就能在本地、免费地转写,还自带时间戳、说话人分离和批量处理。中文尤其强,同时支持 SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种。下面每段代码都已实测。</p><h2>安装</h2><pre><code>pip install -U torch torchaudio@@ -89,7 +89,7 @@ for r in results:<h2>选哪个模型</h2><table><thead><tr><th>model=</th><th>适合</th></tr></thead><tbody>-<tr><td><code>iic/SenseVoiceSmall</code></td><td>默认,极快,50+ 语言,带情感/事件</td></tr>+<tr><td><code>iic/SenseVoiceSmall</code></td><td>默认,极快,SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种,带情感/事件</td></tr><tr><td><code>paraformer-zh</code></td><td>经典中文生产级,稳</td></tr><tr><td><code>FunAudioLLM/Fun-ASR-Nano-2512</code></td><td>旗舰 LLM 解码,中英日 + 中文方言/口音</td></tr><tr><td><code>FunAudioLLM/Fun-ASR-MLT-Nano-2512</code></td><td>独立多语 checkpoint,覆盖 31 语种</td></tr>@@ -100,9 +100,9 @@ for r in results:<ul><li><strong>本地、免费、隐私</strong>:无需联网、无 API key、无时长费用。</li><li><strong>快</strong>:SenseVoice 非自回归,远快于 Whisper(<a href="/blog/funasr-vs-whisper-benchmark.html">实测对比</a>),CPU 也能实时。</li>-<li><strong>中文更准</strong> + 50+ 语言;时间戳、说话人、热词、批量开箱即用。</li>+<li><strong>中文更准</strong> + SenseVoiceSmall 五语种及 Fun-ASR-MLT-Nano 31 语种;时间戳、说话人、热词、批量开箱即用。</li></ul>-<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">在 GitHub 上 Star FunASR ★</a></div>+<div class="cta"><p>FunASR 是通义实验室开源的工业级语音识别工具包。</p><a class="btn" href="/go/github" target="_blank">在 GitHub 上 Star FunASR ★</a></div><!--related-posts--><h2>相关文章</h2><ul class="post-list">@@ -130,5 +130,5 @@ for r in results:<li><a href="/blog/speech-to-text-timestamps-python.html">语音识别带时间戳(字级)</a></li></ul></div></article>-<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · 通义实验室开源语音识别 · <a href="/go/github">GitHub</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/ecosystem.html b/web-pages/product-site/legacy/ecosystem.htmlindex ee439ca21..36a391c4f 100644--- a/web-pages/product-site/legacy/ecosystem.html+++ b/web-pages/product-site/legacy/ecosystem.html@@ -64,7 +64,7 @@ footer a{color:#60a5fa}</style></head><body>-<nav class="nav"><div class="container"><a class="nav-logo" href="./">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="quickstart.html">快速上手</a><a href="models.html">模型</a><a href="ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="en/ecosystem.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a class="nav-logo" href="./">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="quickstart.html">快速上手</a><a href="models.html">模型</a><a href="ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="en/ecosystem.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container">@@ -74,7 +74,7 @@ footer a{color:#60a5fa}<div class="stat-bar"><div class="stat"><div class="stat-num">50+</div><div class="stat-label">集成项目</div></div>-<div class="stat"><div class="stat-num">50+</div><div class="stat-label">支持语种</div></div>+<div class="stat"><div class="stat-num">31</div><div class="stat-label">MLT-Nano 语种</div></div><div class="stat"><div class="stat-num">36K+</div><div class="stat-label">GitHub Stars</div></div><div class="stat"><div class="stat-num">100万+</div><div class="stat-label">月安装量</div></div></div>@@ -412,13 +412,13 @@ funasr-server --device cuda</pre><div class="cta"><p>使用 FunASR 构建了项目?欢迎提交到此列表。</p>-<a class="btn" href="https://github.com/modelscope/FunASR">GitHub</a>+<a class="btn" href="/go/github">GitHub</a><a class="btn outline" href="https://github.com/modelscope/FunASR/issues">提交 Issue</a></div></div></article>-<footer><div class="container">FunASR — 工业级语音识别工具包 — <a href="https://github.com/modelscope/FunASR">GitHub</a> · <a href="./">首页</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer>+<footer><div class="container">FunASR — 工业级语音识别工具包 — <a href="/go/github">GitHub</a> · <a href="./">首页</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/en/blog/funasr-cli-transcribe-command-line.html b/web-pages/product-site/legacy/en/blog/funasr-cli-transcribe-command-line.htmlindex 9b6670a51..576b9af43 100644--- a/web-pages/product-site/legacy/en/blog/funasr-cli-transcribe-command-line.html+++ b/web-pages/product-site/legacy/en/blog/funasr-cli-transcribe-command-line.html@@ -43,7 +43,7 @@ p code,li code{background:var(--surface);padding:2px 6px;border-radius:5px;font-.post-card .date{color:var(--text-muted);font-size:0.8rem}footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-size:0.8rem}footer a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="https://www.funasr.com/blog/funasr-cli-transcribe-command-line.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="https://www.funasr.com/blog/funasr-cli-transcribe-command-line.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Transcribe Audio from the Command Line with FunASR (text / JSON / SRT)</h1><p class="post-meta">FunASR Blog · 2026-06-18 · Tutorial</p>@@ -53,7 +53,7 @@ footer a{color:#94a3b8}</style></head><body>pip install -U funasr # recommended funasr >= 1.3.26</code></pre><h2>Simplest: print text</h2><pre><code>funasr audio.wav</code></pre>-<p>Prints the transcript to the terminal. Defaults to SenseVoice (non-autoregressive, very fast, 50+ languages).</p>+<p>Prints the transcript to the terminal. Defaults to SenseVoice (non-autoregressive, very fast, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages).</p><h2>Generate SRT subtitles</h2><pre><code>funasr audio.wav -f srt -o ./subs</code></pre><p>Writes <code>./subs/audio.srt</code>. Add <code>--spk</code> for speaker-split cues with real timestamps:</p>@@ -95,9 +95,9 @@ print(r.text)</code></pre><ul><li><strong>Local, free, private</strong> - no internet, no API key.</li><li><strong>Fast</strong>: SenseVoice is non-autoregressive, far faster than Whisper (<a href="/en/blog/funasr-vs-whisper-benchmark.html">benchmark</a>) - real-time even on CPU.</li>-<li><strong>Stronger on Chinese</strong> + 50+ languages; subtitles, speakers, hotwords, JSON out of the box.</li>+<li><strong>Stronger on Chinese</strong> + five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages; subtitles, speakers, hotwords, JSON out of the box.</li></ul>-<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">Star FunASR on GitHub ★</a></div>+<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="/go/github" target="_blank">Star FunASR on GitHub ★</a></div><!--related-posts--><h2>Related posts</h2><ul class="post-list">@@ -125,5 +125,5 @@ print(r.text)</code></pre><li><a href="/en/blog/speech-to-text-timestamps-python.html">Speech-to-Text with Timestamps in Python</a></li></ul></div></article>-<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="/go/github">GitHub</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/en/blog/generate-subtitles-srt-vtt-from-audio-video.html b/web-pages/product-site/legacy/en/blog/generate-subtitles-srt-vtt-from-audio-video.htmlindex 56eafb71b..22e1212a3 100644--- a/web-pages/product-site/legacy/en/blog/generate-subtitles-srt-vtt-from-audio-video.html+++ b/web-pages/product-site/legacy/en/blog/generate-subtitles-srt-vtt-from-audio-video.html@@ -1,12 +1,12 @@<!DOCTYPE html><html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">-<title>Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR | FunASR</title><meta name="description" content="Generate subtitles for any video or audio locally and for free: `funasr video.wav -f srt` writes SRT in one line, and Python can also emit VTT with speaker labels and real timestamps. Strong on Chinese, 50+ languages, nothing uploaded. All code tested."><meta name="keywords" content="auto generate subtitles, video to srt, audio to subtitles, srt generator python, generate vtt subtitles, subtitle generator, transcribe video to subtitles, add subtitles to video, speaker labeled subtitles, whisper subtitle alternative">-<link rel="icon" href="/logo.png"><meta property="og:title" content="Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR"><meta property="og:description" content="Generate subtitles for any video or audio locally and for free: `funasr video.wav -f srt` writes SRT in one line, and Python can also emit VTT with speaker labels and real timestamps. Strong on Chinese, 50+ languages, nothing uploaded. All code tested.">+<title>Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR | FunASR</title><meta name="description" content="Generate subtitles for any video or audio locally and for free: `funasr video.wav -f srt` writes SRT in one line, and Python can also emit VTT with speaker labels and real timestamps. Strong on Chinese, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages, nothing uploaded. All code tested."><meta name="keywords" content="auto generate subtitles, video to srt, audio to subtitles, srt generator python, generate vtt subtitles, subtitle generator, transcribe video to subtitles, add subtitles to video, speaker labeled subtitles, whisper subtitle alternative">+<link rel="icon" href="/logo.png"><meta property="og:title" content="Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR"><meta property="og:description" content="Generate subtitles for any video or audio locally and for free: `funasr video.wav -f srt` writes SRT in one line, and Python can also emit VTT with speaker labels and real timestamps. Strong on Chinese, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages, nothing uploaded. All code tested."><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><meta property="og:image" content="https://www.funasr.com/logo.png"><link rel="canonical" href="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link rel="alternate" hreflang="en" href="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link rel="alternate" hreflang="zh" href="https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR","description":"Generate subtitles for any video or audio locally and for free: `funasr video.wav -f srt` writes SRT in one line, and Python can also emit VTT with speaker labels and real timestamps. Strong on Chinese, 50+ languages, nothing uploaded. All code tested.","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR","description":"Generate subtitles for any video or audio locally and for free: `funasr video.wav -f srt` writes SRT in one line, and Python can also emit VTT with speaker labels and real timestamps. Strong on Chinese, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages, nothing uploaded. All code tested.","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -43,11 +43,11 @@ p code,li code{background:var(--surface);padding:2px 6px;border-radius:5px;font-.post-card .date{color:var(--text-muted);font-size:0.8rem}footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-size:0.8rem}footer a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/generate-subtitles-srt-vtt-from-audio-video.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR</h1><p class="post-meta">FunASR Blog · 2026-06-18 · Tutorial</p>-<p>You don't need a cloud service or a paid API to subtitle a video. FunASR turns speech into timestamped <strong>SRT</strong> or <strong>VTT</strong> subtitles locally and for free, and can even label speakers. It's especially strong on Chinese and supports 50+ languages. Every command and snippet below is tested.</p>+<p>You don't need a cloud service or a paid API to subtitle a video. FunASR turns speech into timestamped <strong>SRT</strong> or <strong>VTT</strong> subtitles locally and for free, and can even label speakers. It's especially strong on Chinese and supports five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages. Every command and snippet below is tested.</p><h2>Step 1 (video): extract the audio</h2><p>If your source is a video, pull out 16 kHz mono audio (FunASR's standard input) with ffmpeg:</p>@@ -100,9 +100,9 @@ Speaker 0: 欢迎大家来体验达摩院推出的语音识别模型</code></pre<ul><li><strong>Local, free, private</strong> — no uploads, no API key, no per-minute fees.</li><li><strong>Fast</strong>: SenseVoice is non-autoregressive, far faster than Whisper (<a href="/en/blog/funasr-vs-whisper-benchmark.html">benchmark</a>) — long videos subtitle quickly.</li>-<li><strong>Stronger on Chinese</strong> + 50+ languages; built-in VAD segmentation, speaker labels, and real timestamps.</li>+<li><strong>Stronger on Chinese</strong> + five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages; built-in VAD segmentation, speaker labels, and real timestamps.</li></ul>-<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">Star FunASR on GitHub ★</a></div>+<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="/go/github" target="_blank">Star FunASR on GitHub ★</a></div><!--related-posts--><h2>Related posts</h2><ul class="post-list">@@ -130,5 +130,5 @@ Speaker 0: 欢迎大家来体验达摩院推出的语音识别模型</code></pre<li><a href="/en/blog/speech-to-text-timestamps-python.html">Speech-to-Text with Timestamps in Python</a></li></ul></div></article>-<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="/go/github">GitHub</a></div></footer></body></html>\ No newline at end of filediff --git a/web-pages/product-site/legacy/en/blog/index.html b/web-pages/product-site/legacy/en/blog/index.htmlindex 203ebf217..68641de64 100644--- a/web-pages/product-site/legacy/en/blog/index.html+++ b/web-pages/product-site/legacy/en/blog/index.html@@ -47,8 +47,8 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}@media(max-width:900px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}.launch-feature a{grid-template-columns:1fr}.launch-feature .action{white-space:normal}.previous-release .container{grid-template-columns:1fr}}</style></head><body>-<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="https://www.funasr.com/blog/" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="https://www.funasr.com/blog/" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><section class="launch-feature"><a href="/en/blog/funasr-v1-4-0-pypi-release.html"><div><span class="date">July 31, 2026 · Stable release</span><h2>FunASR v1.4.0: Complete PyPI Packages and Safer AutoModel Arguments</h2><p>SenseVoice and RWKV-BAT package-data fixes, an early vda_model typo guard, and 12 SHA-256-addressed release assets.</p></div><span class="action">Read the release and verification guide →</span></a></section><section class="previous-release"><div class="container"><a class="post-card" href="/en/blog/subtitle-edit-fun-asr-sensevoice-local-subtitles.html"><span class="date">August 3, 2026 · Ecosystem</span><h2>Subtitle Edit 5.2: Local Video Subtitles with Fun-ASR Nano or SenseVoice</h2><p>Select a local ASR engine inside a mature subtitle workspace on Windows, macOS, or Linux, with Q4, Q8, and F16 models.</p></a><a class="post-card" href="/en/blog/funclip-v2-1-0-video-clipping-release.html"><span class="date">July 24, 2026</span><h2>FunClip v2.1.0: The First Versioned Local AI Video-Clipping Release</h2><p>Paraformer, Fun-ASR-Nano, SenseVoice, and LLM clipping in verifiable, rollback-ready source downloads.</p></a></div></section>-<article><div class="container"><h1>FunASR Blog</h1><p class="post-meta">Speech recognition deployment, tutorials and best practices</p><a class="post-card" href="/en/blog/speech-to-text-timestamps-python.html"><span class="date">2026-06-23</span><h2>Speech-to-Text with Word/Character-Level Timestamps in Python</h2><p>FunASR Paraformer gives native character timestamps: every char has [start_ms, end_ms] in one call. Build word highlighting, click-to-seek, subtitle alignment. Real output + pairing code.</p></a><a class="post-card" href="/en/blog/punctuation-restoration-python.html"><span class="date">2026-06-23</span><h2>Punctuation Restoration in Python — Add Punctuation to Text / ASR Output</h2><p>FunASR ct-punc: open-source bilingual (zh+en) punctuation restoration. 3 lines of Python to add ,。? (and capitalize English), or attach to ASR in one line. Real output.</p></a><a class="post-card" href="/en/blog/chinese-speech-recognition.html"><span class="date">2026-06-22</span><h2>Chinese (Mandarin) Speech Recognition in Python — Fast & Accurate with FunASR</h2><p>Purpose-built for Chinese: default flagship Fun-ASR-Nano (CER 8.06%), with SenseVoice (7.81%) / Paraformer (10.18%, timestamps/hotwords) for CPU — all far better than Whisper (~20%).</p></a><a class="post-card" href="/en/blog/self-hosted-google-aws-azure-speech-to-text.html"><span class="date">2026-06-22</span><h2>Self-Hosted Speech-to-Text — Free Alternative to Google / AWS / Azure Cloud Speech APIs</h2><p>Open-source (MIT), local inference, no per-minute billing, audio stays on your network, strong Chinese, OpenAI-compatible (migrate via base_url). Runnable code + FunASR vs cloud comparison + cost analysis.</p></a><a class="post-card" href="/en/blog/voice-activity-detection-python.html"><span class="date">2026-06-22</span><h2>Voice Activity Detection in Python — Detect Speech, Remove Silence, Split Audio</h2><p>FunASR fsmn-vad returns millisecond speech spans in 3 lines. Measured: a 13s clip split into 2 segments in 0.12s, 18% silence removed. Trim silence, split long audio, preprocess Whisper to cut hallucinations.</p></a><a class="post-card" href="/en/blog/japanese-speech-recognition.html"><span class="date">2026-06-22</span><h2>Japanese Speech Recognition in Python — SenseVoice: Transcription + Punctuation + Emotion in One Pass</h2><p>Real side-by-side on one Japanese clip: SenseVoice writes 転売 correctly and adds punctuation; Whisper-small gives 天売 with none. Native ja support + auto language ID + emotion/events, 3 lines of Python.</p></a><a class="post-card" href="/en/blog/self-hosted-deepgram-assemblyai-alternative.html"><span class="date">2026-06-23</span><h2>Self-Hosted Deepgram / AssemblyAI Alternative</h2><p>Open-source, free, self-hosted, audio stays local, Chinese-leading, with an OpenAI-compatible API — replace per-minute cloud STT by changing base_url.</p></a><a class="post-card" href="/en/blog/which-funasr-model.html"><span class="date">2026-06-23</span><h2>Which FunASR Model? Nano vs MLT-Nano vs SenseVoice vs Paraformer</h2><p>A decision table with code: flagship Nano for zh/en/ja plus Chinese dialects/accents, separate MLT-Nano for 31 languages, and CPU choices SenseVoice or Paraformer.</p></a><a class="post-card" href="/en/blog/lightweight-speech-recognition-cpu.html"><span class="date">2026-06-22</span><h2>Lightweight Speech Recognition: Chinese ASR in ~250MB on CPU</h2><p>One binary + a 254MB q8 model, no GPU/Python, 0.16s on CPU, 7.99% CER — smaller than whisper.cpp small and ~3x more accurate.</p></a><a class="post-card" href="/en/blog/cantonese-speech-recognition.html"><span class="date">2026-06-21</span><h2>Cantonese Speech Recognition in Python — SenseVoice Keeps Real Cantonese (Whisper Doesn't)</h2><p>Real side-by-side on one Cantonese clip: SenseVoice keeps 呢/唔/嘅, Whisper rewrites to Mandarin. Native yue support + auto language ID, 3 lines of Python.</p></a><a class="post-card" href="/en/blog/funasr-vs-faster-whisper-chinese.html"><span class="date">2026-06-21</span><h2>FunASR vs faster-whisper: Chinese & Cantonese Compared</h2><p>faster-whisper mislabels Cantonese as Mandarin + Japanese homophone errors; SenseVoice handles Cantonese natively, ~2.7x lower CER on Chinese.</p></a><a class="post-card" href="/en/blog/funasr-llama-cpp-whisper-cpp-alternative.html"><span class="date">2026-06-20</span><h2>FunASR on llama.cpp — a whisper.cpp Alternative for Chinese ASR (CPU, no Python)</h2><p>One self-contained binary, built-in VAD, any audio — download and transcribe Chinese; ~2.7x more accurate than whisper.cpp on CPU.</p></a><a class="post-card" href="/en/blog/speech-to-text-python-transcribe-audio.html"><span class="date">2026-06-18</span><h2>Speech to Text in Python with FunASR</h2><p>Transcribe audio in a few lines of Python — timestamps, speakers, batching. Local, free, no API key.</p></a><a class="post-card" href="/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><span class="date">2026-06-18</span><h2>Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR</h2><p>One command for SRT, Python for VTT too, with speaker labels and real timestamps. Local, free, strong on Chinese.</p></a><a class="post-card" href="/en/blog/self-hosted-openai-whisper-api-alternative.html"><span class="date">2026-06-18</span><h2>Self-Hosted OpenAI Whisper API Alternative with FunASR</h2><p>funasr-server exposes an OpenAI-compatible /v1/audio/transcriptions; the OpenAI SDK works by changing only base_url. Local, free, private.</p></a><a class="post-card" href="/en/blog/funasr-cli-transcribe-command-line.html"><span class="date">2026-06-18</span><h2>Transcribe Audio from the Command Line with FunASR</h2><p>One command -> text/SRT/JSON, --spk for speakers; plus funasr-server for an OpenAI-compatible API.</p></a><a class="post-card" href="/en/blog/funasr-transcribe-long-audio.html"><span class="date">2026-06-17</span><h2>Transcribe Long Audio with FunASR: Hours in One Call</h2><p>Whisper caps at 30s; FunASR ingests any length via built-in VAD - 13 min in 4.3s (186x).</p></a><a class="post-card" href="/en/blog/funasr-realtime-streaming-asr.html"><span class="date">2026-06-17</span><h2>Real-Time Streaming Speech-to-Text with FunASR</h2><p>~600ms low-latency streaming ASR with chunks + cache, plus the 2-pass (streaming + offline) best practice.</p></a><a class="post-card" href="/en/blog/sensevoice-emotion-language-detection.html"><span class="date">2026-06-17</span><h2>Beyond Transcription: Language, Emotion & Audio Events with SenseVoice</h2><p>Transcription + language + emotion + audio events in one pass — the 4-in-1 Whisper cannot do.</p></a><a class="post-card" href="/en/blog/funasr-speaker-diarization.html"><span class="date">2026-06-17</span><h2>Speaker Diarization with FunASR: Who Spoke When</h2><p>Transcription + speaker labels + timestamps in one generate() call. A pyannote+Whisper alternative, no HF gated access.</p></a><a class="post-card" href="/en/blog/funasr-vs-whisper-benchmark.html"><span class="date">2026-06-16</span><h2>FunASR vs Whisper: Real Chinese ASR Benchmark</h2><p>Measured on 184 Chinese files (H100): SenseVoice 169.6x, 7.81% CER — full speed & accuracy data.</p></a><a class="post-card" href="/en/blog/fun-asr-nano-guide.html"><span class="date">2026-06-16</span><h2>Fun-ASR-Nano Guide: 800M End-to-End ASR LLM</h2><p>Flagship for zh/en/ja plus 7 Chinese dialect groups and 26 accents; choose MLT-Nano for 31 languages.</p></a><a class="post-card" href="/en/blog/sensevoice-deployment-guide.html"><span class="date">2026-06-16</span><h2>SenseVoice Deployment Guide: 15x Faster Than Whisper</h2><p>Multilingual ASR in 3 lines, with language/emotion/event tags, VAD, GPU/CPU.</p></a><p>More: <a href="/en/quickstart.html">Quickstart</a> · <a href="/en/models.html">Models</a></p></div></article>+<article><div class="container"><h1>FunASR Blog</h1><p class="post-meta">Speech recognition deployment, tutorials and best practices</p><a class="post-card" href="/en/blog/speech-to-text-timestamps-python.html"><span class="date">2026-06-23</span><h2>Speech-to-Text with Word/Character-Level Timestamps in Python</h2><p>FunASR Paraformer gives native character timestamps: every char has [start_ms, end_ms] in one call. Build word highlighting, click-to-seek, subtitle alignment. Real output + pairing code.</p></a><a class="post-card" href="/en/blog/punctuation-restoration-python.html"><span class="date">2026-06-23</span><h2>Punctuation Restoration in Python — Add Punctuation to Text / ASR Output</h2><p>FunASR ct-punc: open-source bilingual (zh+en) punctuation restoration. 3 lines of Python to add ,。? (and capitalize English), or attach to ASR in one line. Real output.</p></a><a class="post-card" href="/en/blog/chinese-speech-recognition.html"><span class="date">2026-06-22</span><h2>Chinese (Mandarin) Speech Recognition in Python — Fast & Accurate with FunASR</h2><p>Purpose-built for Chinese: default flagship Fun-ASR-Nano (CER 8.06%), with SenseVoice (7.81%) / Paraformer (10.18%, timestamps/hotwords) for CPU — all far better than Whisper (~20%).</p></a><a class="post-card" href="/en/blog/self-hosted-google-aws-azure-speech-to-text.html"><span class="date">2026-06-22</span><h2>Self-Hosted Speech-to-Text — Free Alternative to Google / AWS / Azure Cloud Speech APIs</h2><p>Open-source (MIT), local inference, no per-minute billing, audio stays on your network, strong Chinese, OpenAI-compatible (migrate via base_url). Runnable code + FunASR vs cloud comparison + cost analysis.</p></a><a class="post-card" href="/en/blog/voice-activity-detection-python.html"><span class="date">2026-06-22</span><h2>Voice Activity Detection in Python — Detect Speech, Remove Silence, Split Audio</h2><p>FunASR fsmn-vad returns millisecond speech spans in 3 lines. Measured: a 13s clip split into 2 segments in 0.12s, 18% silence removed. Trim silence, split long audio, preprocess Whisper to cut hallucinations.</p></a><a class="post-card" href="/en/blog/japanese-speech-recognition.html"><span class="date">2026-06-22</span><h2>Japanese Speech Recognition in Python — SenseVoice: Transcription + Punctuation + Emotion in One Pass</h2><p>Real side-by-side on one Japanese clip: SenseVoice writes 転売 correctly and adds punctuation; Whisper-small gives 天売 with none. Native ja support + auto language ID + emotion/events, 3 lines of Python.</p></a><a class="post-card" href="/en/blog/self-hosted-deepgram-assemblyai-alternative.html"><span class="date">2026-06-23</span><h2>Self-Hosted Deepgram / AssemblyAI Alternative</h2><p>Open-source, free, self-hosted, audio stays local, Chinese-leading, with an OpenAI-compatible API — replace per-minute cloud STT by changing base_url.</p></a><a class="post-card" href="/en/blog/which-funasr-model.html"><span class="date">2026-06-23</span><h2>Which FunASR Model? Nano vs MLT-Nano vs SenseVoice vs Paraformer</h2><p>A decision table with code: flagship Nano for zh/en/ja plus Chinese dialects/accents, separate MLT-Nano for 31 languages, and CPU choices SenseVoice or Paraformer.</p></a><a class="post-card" href="/en/blog/lightweight-speech-recognition-cpu.html"><span class="date">2026-06-22</span><h2>Lightweight Speech Recognition: Chinese ASR in ~250MB on CPU</h2><p>One binary + a 254MB q8 model, no GPU/Python, 0.16s on CPU, 7.99% CER — smaller than whisper.cpp small and ~3x more accurate.</p></a><a class="post-card" href="/en/blog/cantonese-speech-recognition.html"><span class="date">2026-06-21</span><h2>Cantonese Speech Recognition in Python — SenseVoice Keeps Real Cantonese (Whisper Doesn't)</h2><p>Real side-by-side on one Cantonese clip: SenseVoice keeps 呢/唔/嘅, Whisper rewrites to Mandarin. Native yue support + auto language ID, 3 lines of Python.</p></a><a class="post-card" href="/en/blog/funasr-vs-faster-whisper-chinese.html"><span class="date">2026-06-21</span><h2>FunASR vs faster-whisper: Chinese & Cantonese Compared</h2><p>faster-whisper mislabels Cantonese as Mandarin + Japanese homophone errors; SenseVoice handles Cantonese natively, ~2.7x lower CER on Chinese.</p></a><a class="post-card" href="/en/blog/funasr-llama-cpp-whisper-cpp-alternative.html"><span class="date">2026-06-20</span><h2>FunASR on llama.cpp — a whisper.cpp Alternative for Chinese ASR (CPU, no Python)</h2><p>One self-contained binary, built-in VAD, any audio — download and transcribe Chinese; ~2.7x more accurate than whisper.cpp on CPU.</p></a><a class="post-card" href="/en/blog/speech-to-text-python-transcribe-audio.html"><span class="date">2026-06-18</span><h2>Speech to Text in Python with FunASR</h2><p>Transcribe audio in a few lines of Python — timestamps, speakers, batching. Local, free, no API key.</p></a><a class="post-card" href="/en/blog/generate-subtitles-srt-vtt-from-audio-video.html"><span class="date">2026-06-18</span><h2>Auto-Generate Subtitles (SRT & VTT) from Audio or Video with FunASR</h2><p>One command for SRT, Python for VTT too, with speaker labels and real timestamps. Local, free, strong on Chinese.</p></a><a class="post-card" href="/en/blog/self-hosted-openai-whisper-api-alternative.html"><span class="date">2026-06-18</span><h2>Self-Hosted OpenAI Whisper API Alternative with FunASR</h2><p>funasr-server exposes an OpenAI-compatible /v1/audio/transcriptions; the OpenAI SDK works by changing only base_url. Local, free, private.</p></a><a class="post-card" href="/en/blog/funasr-cli-transcribe-command-line.html"><span class="date">2026-06-18</span><h2>Transcribe Audio from the Command Line with FunASR</h2><p>One command -> text/SRT/JSON, --spk for speakers; plus funasr-server for an OpenAI-compatible API.</p></a><a class="post-card" href="/en/blog/funasr-transcribe-long-audio.html"><span class="date">2026-06-17</span><h2>Transcribe Long Audio with FunASR: Hours in One Call</h2><p>Whisper caps at 30s; FunASR ingests any length via built-in VAD - 13 min in 4.3s (186x).</p></a><a class="post-card" href="/en/blog/funasr-realtime-streaming-asr.html"><span class="date">2026-06-17</span><h2>Real-Time Streaming Speech-to-Text with FunASR</h2><p>~600ms low-latency streaming ASR with chunks + cache, plus the 2-pass (streaming + offline) best practice.</p></a><a class="post-card" href="/en/blog/sensevoice-emotion-language-detection.html"><span class="date">2026-06-17</span><h2>Beyond Transcription: Language, Emotion & Audio Events with SenseVoice</h2><p>Transcription + language + emotion + audio events in one pass — the 4-in-1 Whisper cannot do.</p></a><a class="post-card" href="/en/blog/funasr-speaker-diarization.html"><span class="date">2026-06-17</span><h2>Speaker Diarization with FunASR: Who Spoke When</h2><p>Transcription + speaker labels + timestamps in one generate() call. A pyannote+Whisper alternative, no HF gated access.</p></a><a class="post-card" href="/en/blog/funasr-vs-whisper-benchmark.html"><span class="date">2026-06-16</span><h2>FunASR vs Whisper: Real Chinese ASR Benchmark</h2><p>Measured on 184 Chinese files (H100): SenseVoice 169.6x, 7.81% CER — full speed & accuracy data.</p></a><a class="post-card" href="/en/blog/fun-asr-nano-guide.html"><span class="date">2026-06-16</span><h2>Fun-ASR-Nano Guide: 800M End-to-End ASR LLM</h2><p>Flagship for zh/en/ja plus 7 Chinese dialect groups and 26 accents; choose MLT-Nano for 31 languages.</p></a><a class="post-card" href="/en/blog/sensevoice-deployment-guide.html"><span class="date">2026-06-16</span><h2>SenseVoice Deployment Guide: Five-Language ASR, Emotion and Audio Events</h2><p>Multilingual ASR in 3 lines, with language/emotion/event tags, VAD, GPU/CPU.</p></a><p>More: <a href="/en/quickstart.html">Quickstart</a> · <a href="/en/models.html">Models</a></p></div></article><footer><div class="container">© 2024-2026 Tongyi Lab, Alibaba Group · <a href="https://github.com/modelscope/FunASR/blob/main/LICENSE">MIT License</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/en/blog/japanese-speech-recognition.html b/web-pages/product-site/legacy/en/blog/japanese-speech-recognition.htmlindex 81284b40f..e5934d037 100644--- a/web-pages/product-site/legacy/en/blog/japanese-speech-recognition.html+++ b/web-pages/product-site/legacy/en/blog/japanese-speech-recognition.html@@ -57,7 +57,7 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/japanese-speech-recognition.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/japanese-speech-recognition.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Japanese Speech Recognition in Python — SenseVoice Does Transcription + Punctuation + Emotion in One Pass (vs Whisper)</h1><p class="post-meta">2026-06-22 · FunASR Team</p>@@ -106,12 +106,12 @@ print(rich_transcription_postprocess(res[0]["text"]))<tr><td>Emotion / audio events</td><td>✅ in one pass</td><td>❌</td></tr><tr><td>Inverse text normalization</td><td>✅ built-in</td><td>partial</td></tr><tr><td>Speed</td><td>non-autoregressive, measured RTF≈0.04 (GPU)</td><td>autoregressive, slower</td></tr>-<tr><td>Multilingual</td><td>one model for 50+ languages (ja/ko/zh/yue/en…)</td><td>multilingual</td></tr>+<tr><td>Multilingual</td><td>one model for Mandarin, Cantonese, English, Japanese, and Korean (ja/ko/zh/yue/en…)</td><td>multilingual</td></tr><tr><td>License</td><td>open-source, commercial-friendly</td><td>open-source</td></tr></table><p>If your use case mixes <strong>Japanese + Chinese + English</strong> (cross-border support, subtitles, meetings), one SenseVoice model covers all of them and bundles emotion analysis. For its emotion/event detection, see <a href="/en/blog/sensevoice-emotion-language-detection.html">SenseVoice emotion & audio event detection</a>; to pick a model, see the <a href="/en/blog/which-funasr-model.html">FunASR model selection guide</a>; and for Chinese accuracy, the <a href="/en/blog/funasr-vs-whisper-benchmark.html">FunASR vs Whisper benchmark</a>.</p>-<div class="cta"><p><strong>The whole FunASR stack is open-source</strong> — industrial-grade ASR / VAD / punctuation / speaker / emotion & events / LLM-ASR, with Japanese and 50+ languages out of the box. If it helps, a GitHub Star really supports the project 👇</p><p><a class="btn" href="https://github.com/QwenAudio/SenseVoice" target="_blank">⭐ Star SenseVoice</a></p><p style="margin-top:10px;font-size:0.85rem">Also star:<a href="https://github.com/modelscope/FunASR" target="_blank">FunASR</a> · <a href="https://github.com/QwenAudio/Fun-ASR" target="_blank">Fun-ASR</a> · <a href="https://github.com/modelscope/FunClip" target="_blank">FunClip</a></p></div>+<div class="cta"><p><strong>The whole FunASR stack is open-source</strong> — industrial-grade ASR / VAD / punctuation / speaker / emotion & events / LLM-ASR, with Japanese and Mandarin, Cantonese, English, Japanese, and Korean out of the box. If it helps, a GitHub Star really supports the project 👇</p><p><a class="btn" href="/go/sensevoice" target="_blank">⭐ Star SenseVoice</a></p><p style="margin-top:10px;font-size:0.85rem">Also star:<a href="/go/github" target="_blank">FunASR</a> · <a href="/go/fun-asr" target="_blank">Fun-ASR</a> · <a href="/go/funclip" target="_blank">FunClip</a></p></div><!--related-posts--><h2>Related posts</h2><ul class="post-list">diff --git a/web-pages/product-site/legacy/en/blog/self-hosted-openai-whisper-api-alternative.html b/web-pages/product-site/legacy/en/blog/self-hosted-openai-whisper-api-alternative.htmlindex e58806d39..1c3a75ef3 100644--- a/web-pages/product-site/legacy/en/blog/self-hosted-openai-whisper-api-alternative.html+++ b/web-pages/product-site/legacy/en/blog/self-hosted-openai-whisper-api-alternative.html@@ -45,7 +45,7 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}@media(max-width:900px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.75rem}table{table-layout:fixed;font-size:0.78rem}th,td{padding:8px 4px;overflow-wrap:anywhere}p code,li code{overflow-wrap:anywhere;word-break:break-word}}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/self-hosted-openai-whisper-api-alternative.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/self-hosted-openai-whisper-api-alternative.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Self-Hosted OpenAI Whisper API Alternative: Drop-in /v1/audio/transcriptions with FunASR</h1><p class="post-meta">FunASR Blog · 2026-06-18 · Tutorial</p>@@ -89,7 +89,7 @@ print(r.text)<h2>Which model to pick</h2><ul>-<li><code>sensevoice</code> — default; non-autoregressive, very fast, 50+ languages, with emotion/events. Best everyday choice.</li>+<li><code>sensevoice</code> — default; non-autoregressive, very fast, Mandarin, Cantonese, English, Japanese, and Korean, with emotion/events. Best everyday choice.</li><li><code>fun-asr-nano</code> — flagship LLM decoder for zh/en/ja plus Chinese dialects/accents (needs vLLM).</li><li>For 31 languages, deploy the separate <code>FunAudioLLM/Fun-ASR-MLT-Nano-2512</code> checkpoint.</li><li><code>paraformer</code> — classic, production-grade Chinese.</li>@@ -119,7 +119,7 @@ AUDIO_STT_OPENAI_API_BASE_URL=http://host.docker.internal:8000/v1AUDIO_STT_OPENAI_API_KEY=funasrAUDIO_STT_MODEL=sensevoice</pre><p>Now Open-WebUI's microphone / voice input runs through FunASR — better Chinese accuracy, data never leaves your machine. The same config works for any tool that supports a custom OpenAI STT endpoint (LibreChat, AnythingLLM, etc.).</p>-<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">Star FunASR on GitHub ★</a></div>+<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="/go/github" target="_blank">Star FunASR on GitHub ★</a></div><!--related-posts--><h2>Related posts</h2><ul class="post-list">@@ -147,5 +147,5 @@ AUDIO_STT_MODEL=sensevoice</pre><li><a href="/en/blog/speech-to-text-timestamps-python.html">Speech-to-Text with Timestamps in Python</a></li></ul></div></article>-<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="/go/github">GitHub</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/en/blog/sensevoice-deployment-guide.html b/web-pages/product-site/legacy/en/blog/sensevoice-deployment-guide.htmlindex 22433516f..5db13738a 100644--- a/web-pages/product-site/legacy/en/blog/sensevoice-deployment-guide.html+++ b/web-pages/product-site/legacy/en/blog/sensevoice-deployment-guide.html@@ -3,12 +3,12 @@<head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">-<title>SenseVoice Deployment Guide: Multilingual ASR 15x Faster Than Whisper | FunASR Blog</title>-<meta name="description" content="Complete SenseVoice deployment tutorial: transcribe multilingual speech in 3 lines of Python, with language ID, emotion and audio-event detection, VAD for long audio, GPU/CPU tips. Open-source ASR 15x faster than Whisper.">+<title>SenseVoice Deployment Guide: Five-Language ASR, Emotion and Audio Events | FunASR Blog</title>+<meta name="description" content="Deploy SenseVoice for Mandarin, Cantonese, English, Japanese, and Korean ASR with language ID, emotion and audio-event detection, long-audio VAD, and GPU/CPU guidance."><meta name="keywords" content="SenseVoice, SenseVoice deployment, speech recognition, multilingual ASR, Whisper alternative, FunASR, emotion recognition, language identification, ASR tutorial, open source speech to text"><link rel="icon" href="/logo.png">-<meta property="og:title" content="SenseVoice Deployment Guide: Multilingual ASR 15x Faster Than Whisper | FunASR Blog">-<meta property="og:description" content="Complete SenseVoice deployment tutorial: transcribe multilingual speech in 3 lines of Python, with language ID, emotion and audio-event detection, VAD for long audio, GPU/CPU tips. Open-source ASR 15x faster than Whisper.">+<meta property="og:title" content="SenseVoice Deployment Guide: Five-Language ASR, Emotion and Audio Events | FunASR Blog">+<meta property="og:description" content="Deploy SenseVoice for Mandarin, Cantonese, English, Japanese, and Korean ASR with language ID, emotion and audio-event detection, long-audio VAD, and GPU/CPU guidance."><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/en/blog/sensevoice-deployment-guide.html"><meta property="og:image" content="https://www.funasr.com/logo.png">@@ -18,7 +18,7 @@<link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/sensevoice-deployment-guide.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"SenseVoice Deployment Guide: Multilingual ASR 15x Faster Than Whisper | FunASR Blog","description":"Complete SenseVoice deployment tutorial: transcribe multilingual speech in 3 lines of Python, with language ID, emotion and audio-event detection, VAD for long audio, GPU/CPU tips. Open-source ASR 15x faster than Whisper.","datePublished":"2026-06-16","dateModified":"2026-06-16","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/sensevoice-deployment-guide.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"SenseVoice Deployment Guide: Five-Language ASR, Emotion and Audio Events | FunASR Blog","description":"Deploy SenseVoice for Mandarin, Cantonese, English, Japanese, and Korean ASR with language ID, emotion and audio-event detection, long-audio VAD, and GPU/CPU guidance.","datePublished":"2026-06-16","dateModified":"2026-08-17","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/sensevoice-deployment-guide.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -54,14 +54,18 @@ p code,li code{background:var(--surface);padding:2px 6px;border-radius:5px;font-.post-card p{margin:0;color:var(--text-soft);font-size:0.9rem}.post-card .date{color:var(--text-muted);font-size:0.8rem}footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-size:0.8rem}-footer a{color:#94a3b8}</style>+footer a{color:#94a3b8}+@media(max-width:700px){+.nav{padding:12px 0}.nav .container{gap:16px}.nav-links{display:none}.nav-btn{white-space:nowrap}+article{padding-top:96px}+}</style></head><body>-<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/sensevoice-deployment-guide.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/sensevoice-deployment-guide.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container">-<h1>SenseVoice Deployment Guide: Multilingual Speech Recognition 15× Faster Than Whisper</h1>+<h1>SenseVoice Deployment Guide: Five-Language ASR, Emotion and Audio Events</h1><p class="post-meta">FunASR Blog · 2026-06-16 · Tutorial</p>-<p><strong>SenseVoice</strong> is an open-source multilingual speech understanding model from the FunAudioLLM team. Beyond automatic speech recognition (ASR), a single forward pass also returns <strong>spoken-language identification, emotion recognition, and audio-event detection</strong>. Its non-autoregressive architecture makes it roughly <strong>15× faster than Whisper-Large</strong>, with latency low enough for both real-time and high-throughput batch workloads.</p>+<p><strong>SenseVoice</strong> is an open-source speech-understanding model from the FunAudioLLM team for Mandarin, Cantonese, English, Japanese, and Korean. Beyond automatic speech recognition (ASR), a single forward pass also returns <strong>spoken-language identification, emotion recognition, and audio-event detection</strong>. Its non-autoregressive architecture suits low-latency and batch workloads; benchmark throughput and latency on the target hardware, audio distribution, and batch configuration.</p><p>Here is a copy-paste guide to get SenseVoice running in three steps.</p><h2>1. Install</h2>@@ -110,8 +114,8 @@ res = model.generate(input="long_audio.wav", language="auto", use_itn=True, batc<table><thead><tr><th>Dimension</th><th>SenseVoice-Small</th></tr></thead><tbody>-<tr><td>Speed vs Whisper-Large</td><td>~15×</td></tr>-<tr><td>Languages</td><td>Chinese / Cantonese / English / Japanese / Korean ...</td></tr>+<tr><td>Performance validation</td><td>Benchmark RTF and p50/p95 on the target hardware, audio distribution, and batch</td></tr>+<tr><td>Languages</td><td>Mandarin / Cantonese / English / Japanese / Korean</td></tr><tr><td>Extra capabilities</td><td>Language ID · Emotion · Audio events</td></tr><tr><td>Architecture</td><td>Non-autoregressive (low latency)</td></tr><tr><td>License</td><td>Open source (commercial use allowed)</td></tr>@@ -121,7 +125,7 @@ res = model.generate(input="long_audio.wav", language="auto", use_itn=True, batc<div class="cta"><h3>Get started with SenseVoice</h3><p>If this guide helped, please star the project on GitHub ⭐ to support open source.</p>-<a href="https://github.com/QwenAudio/SenseVoice" class="btn">SenseVoice GitHub ★</a>+<a href="/go/sensevoice" class="btn">SenseVoice GitHub ★</a></div><p>Read more: <a href="/en/quickstart.html">FunASR Quickstart</a> · <a href="/en/models.html">All models</a> · <a href="/en/ecosystem.html">Ecosystem</a></p>@@ -154,4 +158,4 @@ res = model.generate(input="long_audio.wav", language="auto", use_itn=True, batc</div></article><footer><div class="container">© 2024-2026 Tongyi Lab, Alibaba Group · <a href="https://github.com/modelscope/FunASR/blob/main/LICENSE">MIT License</a></div></footer></body>-</html>\ No newline at end of file+</html>diff --git a/web-pages/product-site/legacy/en/blog/sensevoice-emotion-language-detection.html b/web-pages/product-site/legacy/en/blog/sensevoice-emotion-language-detection.htmlindex a9e0f408b..6dc5e3c62 100644--- a/web-pages/product-site/legacy/en/blog/sensevoice-emotion-language-detection.html+++ b/web-pages/product-site/legacy/en/blog/sensevoice-emotion-language-detection.html@@ -4,11 +4,11 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>Speech Emotion Recognition in Python — Language ID & Audio Events in One Model (SenseVoice) | FunASR Blog</title>-<meta name="description" content="Use SenseVoice to get language ID, emotion (happy/angry/neutral), audio events (music/applause/laughter) and ITN in one inference. 3 lines of Python, real 5-language output, 15x faster than Whisper.">+<meta name="description" content="Use SenseVoice to get language ID, emotion (happy/angry/neutral), audio events (music/applause/laughter) and ITN in one inference. 3 lines of Python, real 5-language output, with throughput that depends on hardware, audio, and batch size."><meta name="keywords" content="speech emotion recognition, speech emotion recognition python, SenseVoice, language identification, audio event detection, multilingual speech recognition, Whisper alternative, FunASR, open source speech to text"><link rel="icon" href="/logo.png"><meta property="og:title" content="Speech Emotion Recognition in Python — Language ID & Audio Events in One Model (SenseVoice) | FunASR Blog">-<meta property="og:description" content="Use SenseVoice to get language ID, emotion (happy/angry/neutral), audio events (music/applause/laughter) and ITN in one inference. 3 lines of Python, real 5-language output, 15x faster than Whisper.">+<meta property="og:description" content="Use SenseVoice to get language ID, emotion (happy/angry/neutral), audio events (music/applause/laughter) and ITN in one inference. 3 lines of Python, real 5-language output, with throughput that depends on hardware, audio, and batch size."><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/en/blog/sensevoice-emotion-language-detection.html"><meta property="og:image" content="https://www.funasr.com/logo.png">@@ -18,7 +18,7 @@<link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/sensevoice-emotion-language-detection.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"Speech Emotion Recognition in Python — Language ID & Audio Events in One Model (SenseVoice) | FunASR Blog","description":"Use SenseVoice to get language ID, emotion (happy/angry/neutral), audio events (music/applause/laughter) and ITN in one inference. 3 lines of Python, real 5-language output, 15x faster than Whisper.","datePublished":"2026-06-19","dateModified":"2026-06-19","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/sensevoice-emotion-language-detection.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"Speech Emotion Recognition in Python — Language ID & Audio Events in One Model (SenseVoice) | FunASR Blog","description":"Use SenseVoice to get language ID, emotion (happy/angry/neutral), audio events (music/applause/laughter) and ITN in one inference. 3 lines of Python, real 5-language output, with throughput that depends on hardware, audio, and batch size.","datePublished":"2026-06-19","dateModified":"2026-06-19","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/sensevoice-emotion-language-detection.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -57,18 +57,18 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}</style></head><body>-<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/sensevoice-emotion-language-detection.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/sensevoice-emotion-language-detection.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Speech Emotion Recognition in Python — Language ID & Audio Events in One Model (SenseVoice)</h1><p class="post-meta">2026-06-19 · FunASR Team</p><p>Most speech-recognition (ASR) models give you only <strong>text</strong>. But real applications often need more: what <strong>emotion</strong> is the speaker in? what <strong>language</strong> is this? is the background <strong>clean speech or music</strong>? With Whisper you have to stack several models — language detection + a separate emotion model + an audio-event classifier. Slow and hard to maintain.</p>-<p><strong>SenseVoice</strong> (an open-source multilingual speech-understanding model from the FunAudioLLM team) returns all of it in <strong>one non-autoregressive forward pass</strong>: the transcript plus <strong>spoken-language ID</strong>, <strong>emotion</strong>, <strong>audio-event detection</strong>, and inverse text normalization (ITN). It is roughly <strong>15× faster than Whisper-Large</strong>.</p>+<p><strong>SenseVoice</strong> (an open-source multilingual speech-understanding model from the FunAudioLLM team) returns all of it in <strong>one non-autoregressive forward pass</strong>: the transcript plus <strong>spoken-language ID</strong>, <strong>emotion</strong>, <strong>audio-event detection</strong>, and inverse text normalization (ITN). Its non-autoregressive architecture suits low-latency and batch inference; benchmark throughput on the target hardware and audio distribution.</p><h2>What one inference gives you</h2><table><tr><th>Capability</th><th>Detail</th></tr>-<tr><td>ASR</td><td>50+ languages, leading Chinese accuracy</td></tr>+<tr><td>ASR</td><td>Mandarin, Cantonese, English, Japanese, and Korean, leading Chinese accuracy</td></tr><tr><td>Language ID</td><td>auto-detects <code>zh / en / ja / ko / yue …</code></td></tr><tr><td>Emotion</td><td>HAPPY 😊 / SAD 😔 / ANGRY 😡 / NEUTRAL / FEARFUL / DISGUSTED / SURPRISED</td></tr><tr><td>Audio events</td><td>Speech / BGM 🎵 / Applause 👏 / Laughter / Cry</td></tr>@@ -125,7 +125,7 @@ print(language, emotion, event, text)</pre></table><p>When you need <strong>understanding</strong> of audio and not just text, one SenseVoice model replaces a language-detector + emotion-model + event-classifier stack.</p>-<div class="cta"><p><strong>The whole FunASR stack is open-source</strong> — industrial-grade ASR / VAD / punctuation / speaker / emotion & events / LLM-ASR. If it helps, a GitHub Star really supports the project 👇</p><p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">⭐ Star FunASR</a></p><p style="margin-top:10px;font-size:0.85rem">Also star:<a href="https://github.com/QwenAudio/SenseVoice" target="_blank">SenseVoice</a> · <a href="https://github.com/QwenAudio/Fun-ASR" target="_blank">Fun-ASR</a> · <a href="https://github.com/modelscope/FunClip" target="_blank">FunClip</a></p></div>+<div class="cta"><p><strong>The whole FunASR stack is open-source</strong> — industrial-grade ASR / VAD / punctuation / speaker / emotion & events / LLM-ASR. If it helps, a GitHub Star really supports the project 👇</p><p><a class="btn" href="/go/github" target="_blank">⭐ Star FunASR</a></p><p style="margin-top:10px;font-size:0.85rem">Also star:<a href="/go/sensevoice" target="_blank">SenseVoice</a> · <a href="/go/fun-asr" target="_blank">Fun-ASR</a> · <a href="/go/funclip" target="_blank">FunClip</a></p></div><!--related-posts--><h2>Related posts</h2><ul class="post-list">diff --git a/web-pages/product-site/legacy/en/blog/speech-to-text-python-transcribe-audio.html b/web-pages/product-site/legacy/en/blog/speech-to-text-python-transcribe-audio.htmlindex 55bb18c24..da31bfbfa 100644--- a/web-pages/product-site/legacy/en/blog/speech-to-text-python-transcribe-audio.html+++ b/web-pages/product-site/legacy/en/blog/speech-to-text-python-transcribe-audio.html@@ -1,12 +1,12 @@<!DOCTYPE html><html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">-<title>Speech to Text in Python: Transcribe Audio Locally and Free with FunASR | FunASR</title><meta name="description" content="Turn audio into text with a few lines of Python: `AutoModel(...).generate(...)` — local, free, no API key. Timestamps, speakers, batch processing; strong on Chinese, 50+ languages. All code tested."><meta name="keywords" content="python speech to text, transcribe audio python, speech recognition python, audio to text python, python transcription, offline speech to text python, free speech to text python, funasr python">-<link rel="icon" href="/logo.png"><meta property="og:title" content="Speech to Text in Python: Transcribe Audio Locally and Free with FunASR"><meta property="og:description" content="Turn audio into text with a few lines of Python: `AutoModel(...).generate(...)` — local, free, no API key. Timestamps, speakers, batch processing; strong on Chinese, 50+ languages. All code tested.">+<title>Speech to Text in Python: Transcribe Audio Locally and Free with FunASR | FunASR</title><meta name="description" content="Turn audio into text with a few lines of Python: `AutoModel(...).generate(...)` — local, free, no API key. Timestamps, speakers, batch processing; strong on Chinese, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages. All code tested."><meta name="keywords" content="python speech to text, transcribe audio python, speech recognition python, audio to text python, python transcription, offline speech to text python, free speech to text python, funasr python">+<link rel="icon" href="/logo.png"><meta property="og:title" content="Speech to Text in Python: Transcribe Audio Locally and Free with FunASR"><meta property="og:description" content="Turn audio into text with a few lines of Python: `AutoModel(...).generate(...)` — local, free, no API key. Timestamps, speakers, batch processing; strong on Chinese, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages. All code tested."><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"><meta property="og:image" content="https://www.funasr.com/logo.png"><link rel="canonical" href="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"><link rel="alternate" hreflang="en" href="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"><link rel="alternate" hreflang="zh" href="https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html"><link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"BlogPosting","headline":"Speech to Text in Python: Transcribe Audio Locally and Free with FunASR","description":"Turn audio into text with a few lines of Python: `AutoModel(...).generate(...)` — local, free, no API key. Timestamps, speakers, batch processing; strong on Chinese, 50+ languages. All code tested.","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"},"image":"https://www.funasr.com/logo.png"}+{"@context":"https://schema.org","@type":"BlogPosting","headline":"Speech to Text in Python: Transcribe Audio Locally and Free with FunASR","description":"Turn audio into text with a few lines of Python: `AutoModel(...).generate(...)` — local, free, no API key. Timestamps, speakers, batch processing; strong on Chinese, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages. All code tested.","datePublished":"2026-06-18","dateModified":"2026-06-18","author":{"@type":"Organization","name":"FunASR / Tongyi Lab"},"publisher":{"@type":"Organization","name":"FunASR","logo":{"@type":"ImageObject","url":"https://www.funasr.com/logo.png"}},"mainEntityOfPage":{"@type":"WebPage","@id":"https://www.funasr.com/en/blog/speech-to-text-python-transcribe-audio.html"},"image":"https://www.funasr.com/logo.png"}</script><style>:root{--primary:#2563eb;--primary-dark:#1d4ed8;--bg:#fff;--surface:#f8fafc;--border:#e2e8f0;--text:#0f172a;--text-soft:#475569;--text-muted:#94a3b8;--radius:12px}*{margin:0;padding:0;box-sizing:border-box}@@ -45,11 +45,11 @@ footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-sifooter a{color:#94a3b8}@media(max-width:900px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.75rem}table{table-layout:fixed;font-size:0.78rem}th,td{padding:8px 4px;overflow-wrap:anywhere}p code,li code{overflow-wrap:anywhere;word-break:break-word}}</style></head><body>-<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Donors</a></div><a href="https://www.funasr.com/blog/speech-to-text-python-transcribe-audio.html" style="font-size:0.83rem;font-weight:600;color:var(--text-soft)">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container"><h1>Speech to Text in Python: Transcribe Audio Locally and Free with FunASR</h1><p class="post-meta">FunASR Blog · 2026-06-18 · Tutorial</p>-<p>You don't need a cloud API or a heavy Whisper setup to transcribe audio in Python. FunASR turns audio into text in a few lines, locally and for free, with timestamps, speaker diarization, and batch processing built in. It's especially strong on Chinese and supports 50+ languages. Every snippet below is tested.</p>+<p>You don't need a cloud API or a heavy Whisper setup to transcribe audio in Python. FunASR turns audio into text in a few lines, locally and for free, with timestamps, speaker diarization, and batch processing built in. It's especially strong on Chinese and supports five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages. Every snippet below is tested.</p><h2>Install</h2><pre><code>pip install -U torch torchaudio@@ -89,7 +89,7 @@ for r in results:<h2>Which model to pick</h2><table><thead><tr><th>model=</th><th>Best for</th></tr></thead><tbody>-<tr><td><code>iic/SenseVoiceSmall</code></td><td>Default; very fast, 50+ languages, emotion/events</td></tr>+<tr><td><code>iic/SenseVoiceSmall</code></td><td>Default; very fast, five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages, emotion/events</td></tr><tr><td><code>paraformer-zh</code></td><td>Classic, production-grade Chinese</td></tr><tr><td><code>FunAudioLLM/Fun-ASR-Nano-2512</code></td><td>Flagship LLM decoder for zh/en/ja + Chinese dialects/accents</td></tr><tr><td><code>FunAudioLLM/Fun-ASR-MLT-Nano-2512</code></td><td>Separate multilingual checkpoint covering 31 languages</td></tr>@@ -100,9 +100,9 @@ for r in results:<ul><li><strong>Local, free, private</strong> — no internet, no API key, no per-minute fees.</li><li><strong>Fast</strong>: SenseVoice is non-autoregressive, far faster than Whisper (<a href="/en/blog/funasr-vs-whisper-benchmark.html">benchmark</a>) — real-time even on CPU.</li>-<li><strong>Stronger on Chinese</strong> + 50+ languages; timestamps, speakers, hotwords, and batching out of the box.</li>+<li><strong>Stronger on Chinese</strong> + five SenseVoiceSmall languages and 31 Fun-ASR-MLT-Nano languages; timestamps, speakers, hotwords, and batching out of the box.</li></ul>-<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="https://github.com/modelscope/FunASR" target="_blank">Star FunASR on GitHub ★</a></div>+<div class="cta"><p>FunASR is Tongyi Lab's open-source, industrial-grade speech recognition toolkit.</p><a class="btn" href="/go/github" target="_blank">Star FunASR on GitHub ★</a></div><!--related-posts--><h2>Related posts</h2><ul class="post-list">@@ -130,5 +130,5 @@ for r in results:<li><a href="/en/blog/speech-to-text-timestamps-python.html">Speech-to-Text with Timestamps in Python</a></li></ul></div></article>-<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="https://github.com/modelscope/FunASR">GitHub</a></div></footer>+<footer><div class="container">FunASR · Open-source speech recognition by Tongyi Lab · <a href="/go/github">GitHub</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/en/ecosystem.html b/web-pages/product-site/legacy/en/ecosystem.htmlindex 390e4687e..3cda82775 100644--- a/web-pages/product-site/legacy/en/ecosystem.html+++ b/web-pages/product-site/legacy/en/ecosystem.html@@ -65,7 +65,7 @@ footer a{color:#60a5fa}</style></head><body>-<nav class="nav"><div class="container"><a class="nav-logo" href="../en/">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="quickstart.html">Quickstart</a><a href="models.html">Models</a><a href="ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="/ecosystem.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></nav>+<nav class="nav"><div class="container"><a class="nav-logo" href="../en/">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="quickstart.html">Quickstart</a><a href="models.html">Models</a><a href="ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="/ecosystem.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container">@@ -75,7 +75,7 @@ footer a{color:#60a5fa}<div class="stat-bar"><div class="stat"><div class="stat-num">50+</div><div class="stat-label">Integrations</div></div>-<div class="stat"><div class="stat-num">50+</div><div class="stat-label">Languages</div></div>+<div class="stat"><div class="stat-num">31</div><div class="stat-label">MLT-Nano Languages</div></div><div class="stat"><div class="stat-num">36K+</div><div class="stat-label">GitHub Stars</div></div><div class="stat"><div class="stat-num">1M+</div><div class="stat-label">pip installs/month</div></div></div>@@ -415,13 +415,13 @@ funasr-server --device cuda</pre><div class="cta"><p>Build something with FunASR? We'd love to list it here.</p>-<a class="btn" href="https://github.com/modelscope/FunASR">GitHub</a>+<a class="btn" href="/go/github">GitHub</a><a class="btn outline" href="https://github.com/modelscope/FunASR/issues">Open an Issue</a></div></div></article>-<footer><div class="container">FunASR — Industrial-grade speech recognition toolkit — <a href="https://github.com/modelscope/FunASR">GitHub</a> · <a href="../en/">Home</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer>+<footer><div class="container">FunASR — Industrial-grade speech recognition toolkit — <a href="/go/github">GitHub</a> · <a href="../en/">Home</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/en/index.html b/web-pages/product-site/legacy/en/index.htmlindex b176644f4..33d8ca97f 100644--- a/web-pages/product-site/legacy/en/index.html+++ b/web-pages/product-site/legacy/en/index.html@@ -4,7 +4,7 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>FunASR - Industrial Speech Recognition | All-in-One Speech Understanding</title>-<meta name="description" content="FunASR is an industrial-grade speech recognition service offering ASR, voice activity detection, punctuation restoration, speaker diarization, emotion detection, and audio event recognition through a unified Python API. 50+ languages, self-hosted.">+<meta name="description" content="FunASR is an industrial-grade speech recognition service offering ASR, voice activity detection, punctuation restoration, speaker diarization, emotion detection, and audio event recognition through a unified Python API, with model-specific language coverage and self-hosted deployment."><link rel="icon" href="../logo.png"><meta property="og:title" content="FunASR - All-in-One Industrial Speech Understanding"><meta property="og:description" content="Speech recognition, VAD, punctuation, speaker diarization, emotion detection, audio events — one unified Python API, deploy with one command.">@@ -119,7 +119,7 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}@media(max-width:600px){.hero-flagship{display:block;width:100%;padding:7px 12px;font-size:0.78rem;border-radius:10px}}</style><script type="application/ld+json">-{"@context":"https://schema.org","@type":"SoftwareApplication","name":"FunASR","applicationCategory":"DeveloperApplication","operatingSystem":"Linux, Windows, macOS","offers":{"@type":"Offer","price":"0","priceCurrency":"USD"},"description":"Industrial-grade speech recognition toolkit supporting 50+ languages","url":"https://www.funasr.com","downloadUrl":"https://pypi.org/project/funasr/","author":{"@type":"Organization","name":"FunAudioLLM, Tongyi Lab"}}+{"@context":"https://schema.org","@type":"SoftwareApplication","name":"FunASR","applicationCategory":"DeveloperApplication","operatingSystem":"Linux, Windows, macOS","offers":{"@type":"Offer","price":"0","priceCurrency":"USD"},"description":"Industrial-grade speech recognition toolkit with model-specific language coverage and self-hosted deployment","url":"https://www.funasr.com","downloadUrl":"https://pypi.org/project/funasr/","author":{"@type":"Organization","name":"FunAudioLLM, Tongyi Lab"}}</script></head><body>@@ -139,7 +139,7 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}</div><a href="/" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">中文</a>-<a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a>+<a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><!-- Hero -->@@ -151,14 +151,14 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}<a href="https://huggingface.co/FunAudioLLM/Fun-ASR-Nano-2512" target="_blank" rel="noopener" class="hero-flagship">🚀 Flagship <strong>Fun-ASR-Nano</strong> · zh/en/ja + Chinese dialects/accents</a><a href="https://huggingface.co/FunAudioLLM/Fun-ASR-MLT-Nano-2512" target="_blank" rel="noopener" class="hero-flagship mlt">🌍 31 languages <strong>Fun-ASR-MLT-Nano</strong> · separate checkpoint</a></div>-<p class="hero-sub">Speech recognition, voice activity detection, punctuation restoration, speaker diarization, emotion detection, and audio event recognition — one unified Python API handles it all. 50+ languages, self-hosted, production-ready.</p>+<p class="hero-sub">Speech recognition, voice activity detection, punctuation restoration, speaker diarization, emotion detection, and audio event recognition — one unified Python API handles it all. Language coverage is model-specific, with self-hosted production deployment.</p><div class="hero-actions"><a href="#usage" class="btn-primary">How to Use</a><a href="#demo" class="btn-outline">Watch Demo</a></div><div class="hero-stats">-<div class="hero-stat"><strong>50+</strong><span>Languages</span></div>-<div class="hero-stat"><strong>170x</strong><span>Realtime Speed</span></div>+<div class="hero-stat"><strong>31</strong><span>MLT-Nano Languages</span></div>+<div class="hero-stat"><strong>5</strong><span>SenseVoice Languages</span></div><div class="hero-stat"><strong>1 API</strong><span>Unified Interface</span></div></div></div>@@ -187,7 +187,7 @@ print(result.text)<p>A complete speech understanding pipeline from raw audio to structured output — all in one call</p></div><div class="capabilities-grid">-<div class="cap-card"><div class="cap-icon">🎙️</div><h3>Speech Recognition</h3><p>End-to-end ASR supporting 50+ languages, including 7 Chinese dialects and 26 regional accents, with automatic language detection</p></div>+<div class="cap-card"><div class="cap-icon">🎙️</div><h3>Speech Recognition</h3><p>End-to-end ASR: SenseVoiceSmall covers Mandarin, Cantonese, English, Japanese, and Korean; MLT-Nano covers 31 languages; Nano targets Chinese dialects and accents</p></div><div class="cap-card"><div class="cap-icon">📍</div><h3>Voice Activity Detection</h3><p>Millisecond-precision VAD with adaptive silence thresholds, accurately segmenting speech from silence</p></div><div class="cap-card"><div class="cap-icon">✍️</div><h3>Punctuation Restoration</h3><p>Automatically adds punctuation and applies inverse text normalization, producing readable formatted text</p></div><div class="cap-card"><div class="cap-icon">👥</div><h3>Speaker Diarization</h3><p>Identifies who said what, labeling each sentence with a speaker ID — ideal for meetings and interviews</p></div>@@ -297,7 +297,7 @@ print(result.text)<div class="more-grid"><a class="more-item" href="https://github.com/modelscope/FunASR/blob/main/docs/vllm_guide.md" target="_blank">🔥 Hotword Boosting <span>→</span></a><a class="more-item" href="https://modelscope.github.io/FunASR/tutorial.html" target="_blank">👥 Speaker Diarization <span>→</span></a>-<a class="more-item" href="https://github.com/QwenAudio/SenseVoice" target="_blank">😊 Emotion Detection <span>→</span></a>+<a class="more-item" href="/go/sensevoice" target="_blank">😊 Emotion Detection <span>→</span></a><a class="more-item" href="https://modelscope.github.io/FunASR/tutorial.html" target="_blank">🌍 Multilingual <span>→</span></a><a class="more-item" href="https://modelscope.github.io/FunASR/training.html" target="_blank">🎯 Fine-tuning <span>→</span></a><a class="more-item" href="https://github.com/modelscope/FunASR/blob/main/docs/vllm_guide.md" target="_blank">⚡ vLLM Acceleration <span>→</span></a>@@ -348,13 +348,13 @@ print(result.text)<a href="https://modelscope.github.io/FunASR/api.html">API Reference</a></div><div class="footer-col"><h4>Ecosystem</h4>-<a href="https://github.com/QwenAudio/Fun-ASR">Fun-ASR-Nano</a>-<a href="https://github.com/QwenAudio/SenseVoice">SenseVoice</a>-<a href="https://github.com/modelscope/FunClip">FunClip</a>+<a href="/go/fun-asr">Fun-ASR-Nano</a>+<a href="/go/sensevoice">SenseVoice</a>+<a href="/go/funclip">FunClip</a><a href="https://github.com/QwenAudio/CosyVoice">CosyVoice</a></div><div class="footer-col"><h4>Community</h4>-<a href="https://github.com/modelscope/FunASR">GitHub</a>+<a href="/go/github">GitHub</a><a href="https://github.com/modelscope/FunASR/discussions">Discussions</a><a href="https://github.com/modelscope/FunASR/issues">Issues</a></div>diff --git a/web-pages/product-site/legacy/en/models.html b/web-pages/product-site/legacy/en/models.htmlindex 0472a07db..ba543ed47 100644--- a/web-pages/product-site/legacy/en/models.html+++ b/web-pages/product-site/legacy/en/models.html@@ -4,7 +4,7 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>FunASR Models — SenseVoice, Paraformer, Fun-ASR-Nano, cam++</title>-<meta name="description" content="Choose the right FunASR model: SenseVoice (170x realtime, 50+ langs), Paraformer (Chinese accuracy and reliable character timestamps), Fun-ASR-Nano (LLM-based, vLLM accelerated), and cam++ (speaker diarization).">+<meta name="description" content="Choose the right FunASR model: SenseVoice (Mandarin, Cantonese, English, Japanese, and Korean), Paraformer (Chinese accuracy and reliable character timestamps), Fun-ASR-Nano (LLM-based, vLLM accelerated), and cam++ (speaker diarization)."><link rel="icon" href="../logo.png"><meta property="og:title" content="FunASR Models — Speech Recognition Model Comparison"><meta property="og:description" content="Compare FunASR models: SenseVoice, Paraformer, Fun-ASR-Nano for speech recognition. Choose by speed, accuracy, language support, and features.">@@ -60,7 +60,7 @@ footer a{color:#60a5fa}</style></head><body>-<nav class="nav"><div class="container"><a class="nav-logo" href="../en/">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="quickstart.html">Quickstart</a><a href="models.html">Models</a><a href="ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a><a href="/models.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">中文</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></div></nav>+<nav class="nav"><div class="container"><a class="nav-logo" href="../en/">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="quickstart.html">Quickstart</a><a href="models.html">Models</a><a href="ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a><a href="/models.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">中文</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></div></nav><article><div class="container">@@ -75,10 +75,10 @@ footer a{color:#60a5fa}<tr><th>Model</th><th>Speed</th><th>Languages</th><th>Params</th><th>Best For</th></tr></thead><tbody>-<tr><td><strong>Fun-ASR-Nano</strong> ⭐</td><td>vLLM 340x</td><td>zh/en/ja + Chinese dialects/accents</td><td>800M</td><td>Flagship · LLM-ASR · hardest cases</td></tr>+<tr><td><strong>Fun-ASR-Nano</strong> ⭐</td><td>vLLM; benchmark your hardware</td><td>zh/en/ja + Chinese dialects/accents</td><td>800M</td><td>Flagship · LLM-ASR · hardest cases</td></tr><tr><td><strong>Fun-ASR-MLT-Nano</strong></td><td>vLLM</td><td>31</td><td>800M</td><td>Separate multilingual checkpoint</td></tr>-<tr><td><strong>SenseVoice Small</strong></td><td>170x realtime</td><td>50+</td><td>234M</td><td>Fast multilingual, emotion detection</td></tr>-<tr><td><strong>Paraformer-zh</strong></td><td>120x realtime</td><td>zh, yue</td><td>220M</td><td>Best Chinese accuracy</td></tr>+<tr><td><strong>SenseVoice Small</strong></td><td>Non-autoregressive; benchmark your hardware</td><td>zh, yue, en, ja, ko</td><td>234M</td><td>Five-language ASR, emotion detection</td></tr>+<tr><td><strong>Paraformer-zh</strong></td><td>Non-autoregressive; benchmark your hardware</td><td>zh, yue</td><td>220M</td><td>Best Chinese accuracy</td></tr><tr><td><strong>cam++</strong></td><td>realtime</td><td>any</td><td>7.2M</td><td>Speaker diarization & verification</td></tr></tbody></table>@@ -123,10 +123,10 @@ result = model.generate(input="audio.wav")</pre><div class="model-card"><div class="model-name">SenseVoice Small</div><div class="model-meta">234M params · Non-autoregressive · <a href="/go/sensevoice">GitHub</a> · <a href="https://huggingface.co/FunAudioLLM/SenseVoiceSmall">HuggingFace</a></div>-<div class="model-desc">Ultra-fast speech recognition with built-in emotion and audio event detection. Supports 50+ languages including Chinese, English, Japanese, Korean, French, German, and more. Non-autoregressive architecture delivers 170x realtime speed on GPU.</div>+<div class="model-desc">Ultra-fast speech recognition with built-in emotion and audio event detection. Supports Mandarin, Cantonese, English, Japanese, and Korean. Its non-autoregressive architecture suits low-latency deployment; throughput depends on hardware, audio length, and batch configuration.</div><div class="model-features">-<span class="badge badge-green">170x realtime</span>-<span class="badge badge-blue">50+ languages</span>+<span class="badge badge-green">Non-autoregressive</span>+<span class="badge badge-blue">Five languages</span><span class="badge badge-orange">Emotion detection</span><span class="badge badge-blue">Audio events</span><span class="badge badge-green">CPU-viable</span>@@ -144,7 +144,7 @@ print(result[0]["text"])</pre><div class="model-meta">220M params · Non-autoregressive · <a href="https://huggingface.co/funasr/paraformer-zh">HuggingFace</a></div><div class="model-desc">Highest-accuracy Chinese speech recognition model. Non-autoregressive with CTC-guided attention, trained on 60,000+ hours of Mandarin speech. Includes built-in punctuation restoration and timestamp prediction.</div><div class="model-features">-<span class="badge badge-green">120x realtime</span>+<span class="badge badge-green">Non-autoregressive; benchmark your hardware</span><span class="badge badge-blue">Chinese + Cantonese</span><span class="badge badge-green">Best accuracy</span><span class="badge badge-blue">Timestamps</span>@@ -230,6 +230,6 @@ curl http://localhost:8000/v1/audio/transcriptions \</div></article>-<footer><div class="container">FunASR — Industrial-grade speech recognition toolkit — <a href="https://github.com/modelscope/FunASR">GitHub</a> · <a href="../en/">Home</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer>+<footer><div class="container">FunASR — Industrial-grade speech recognition toolkit — <a href="/go/github">GitHub</a> · <a href="../en/">Home</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/en/quickstart.html b/web-pages/product-site/legacy/en/quickstart.htmlindex 6d8de4fd5..bfa8e53ec 100644--- a/web-pages/product-site/legacy/en/quickstart.html+++ b/web-pages/product-site/legacy/en/quickstart.html@@ -41,7 +41,7 @@"@type": "HowToStep","position": 2,"name": "Transcribe Audio Files",- "text": "Use AutoModel to load the SenseVoice model and call the generate method to transcribe audio files. Supports 50+ languages with emotion detection.",+ "text": "Use AutoModel to load the SenseVoice model and call the generate method to transcribe audio files. SenseVoiceSmall supports Mandarin, Cantonese, English, Japanese, and Korean with emotion detection.","url": "https://www.funasr.com/en/quickstart.html#file-transcription"},{@@ -141,7 +141,7 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}<a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="quickstart.html">Quickstart</a><a href="models.html">Models</a><a href="ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="/quickstart.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">中文</a>-<a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a>+<a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><!-- Hero -->@@ -178,7 +178,7 @@ If you just need a one-off transcription, <code>pip install funasr</code> is eno<section class="alt" id="file-transcription"><div class="container"><div class="guide-section"><h2><div class="step-number">1</div>File Transcription</h2>-<p class="subtitle">Pass in an audio file, get back the full transcript. Supports 50+ languages with built-in emotion and audio event detection.</p>+<p class="subtitle">Pass in an audio file, get back the full transcript. SenseVoiceSmall supports Mandarin, Cantonese, English, Japanese, and Korean with built-in emotion and audio event detection.</p><div class="code-block"><span class="k">from</span> funasr <span class="k">import</span> AutoModel<br>@@ -254,7 +254,7 @@ Let's discuss the...three topics today. Sounds good. First one is...the Q3 plan.<div class="tip-block"><strong>WebSocket Real-time Service</strong>-Need browser-based live transcription? FunASR ships with a complete WebSocket server solution. Check the <a href="https://github.com/modelscope/FunASR" target="_blank">GitHub repository</a> under <code>examples/</code> for the WebSocket server and browser client code.+Need browser-based live transcription? FunASR ships with a complete WebSocket server solution. Check the <a href="/go/github" target="_blank">GitHub repository</a> under <code>examples/</code> for the WebSocket server and browser client code.</div></div></div></section>@@ -328,7 +328,7 @@ MCP (Model Context Protocol) is a standard protocol for AI assistants to invoke<p>Explore more of what FunASR can do</p></div><div class="next-grid">-<a class="next-card" href="https://github.com/modelscope/FunASR" target="_blank">+<a class="next-card" href="/go/github" target="_blank"><div class="next-icon">⭐</div><div><h3>GitHub Repository</h3><p>Source code, examples, and issue tracker</p></div></a>@@ -361,13 +361,13 @@ MCP (Model Context Protocol) is a standard protocol for AI assistants to invoke<a href="https://modelscope.github.io/FunASR/api.html">API Reference</a></div><div class="footer-col"><h4>Ecosystem</h4>-<a href="https://github.com/QwenAudio/Fun-ASR">Fun-ASR-Nano</a>-<a href="https://github.com/QwenAudio/SenseVoice">SenseVoice</a>-<a href="https://github.com/modelscope/FunClip">FunClip</a>+<a href="/go/fun-asr">Fun-ASR-Nano</a>+<a href="/go/sensevoice">SenseVoice</a>+<a href="/go/funclip">FunClip</a><a href="https://github.com/QwenAudio/CosyVoice">CosyVoice</a></div><div class="footer-col"><h4>Community</h4>-<a href="https://github.com/modelscope/FunASR">GitHub</a>+<a href="/go/github">GitHub</a><a href="https://github.com/modelscope/FunASR/discussions">Discussions</a><a href="https://github.com/modelscope/FunASR/issues">Issues</a></div>diff --git a/web-pages/product-site/legacy/en/vs-whisper.html b/web-pages/product-site/legacy/en/vs-whisper.htmlindex 741437468..e86e41cfe 100644--- a/web-pages/product-site/legacy/en/vs-whisper.html+++ b/web-pages/product-site/legacy/en/vs-whisper.html@@ -4,10 +4,10 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>FunASR vs Whisper — Open Source Speech Recognition Comparison 2026</title>-<meta name="description" content="Compare FunASR and Whisper: speed, accuracy, features, languages. FunASR is 170x realtime vs Whisper's 13x. Built-in speaker diarization, emotion detection, streaming.">+<meta name="description" content="Compare FunASR and Whisper by model coverage, diarization, emotion detection, streaming, CPU/GPU deployment, and reproducible benchmarking."><link rel="icon" href="../logo.png"><meta property="og:title" content="FunASR vs Whisper — Which Open Source ASR Should You Use?">-<meta property="og:description" content="Side-by-side comparison: FunASR runs 13x faster than Whisper on GPU, works on CPU, includes speaker ID and emotion detection out of the box.">+<meta property="og:description" content="Compare FunASR and Whisper by model coverage, speech features, runtime options, and target-hardware benchmarks."><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/en/vs-whisper.html"><meta property="og:image" content="https://www.funasr.com/logo.png">@@ -55,7 +55,7 @@ code{font-family:'JetBrains Mono',monospace}.badge{display:inline-block;padding:2px 8px;background:#dcfce7;color:#166534;border-radius:4px;font-size:0.75rem;font-weight:600}footer{background:#0f172a;color:#94a3b8;padding:32px 0;text-align:center;font-size:0.82rem}footer a{color:#94a3b8}-@media(max-width:600px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.7rem}table{font-size:0.8rem}th,td{padding:8px}}+@media(max-width:600px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.7rem}table{font-size:0.8rem;table-layout:fixed}th,td{padding:8px;overflow-wrap:anywhere}}</style></head><body>@@ -63,7 +63,7 @@ footer a{color:#94a3b8}<a href="/en/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/en/#capabilities">Capabilities</a><a href="/en/#usage">How to Use</a><a href="/en/#benchmark">Benchmark</a><a href="/en/#demo">Demo</a><a href="https://modelscope.github.io/FunASR/" target="_blank">Docs</a><a href="/en/quickstart.html">Quickstart</a><a href="/en/models.html">Models</a><a href="/en/ecosystem.html">Ecosystem</a><a href="/en/blog/">Blog</a><a href="/en/vs-whisper.html">vs Whisper</a><a href="/en/donors.html">Thanks</a></div><a href="/" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">中文</a>-<a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a>+<a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container">@@ -71,28 +71,24 @@ footer a{color:#94a3b8}<h1>FunASR vs Whisper: Which Open Source ASR Should You Use?</h1><p>Both FunASR and OpenAI Whisper are open-source speech recognition tools. Here's a detailed comparison to help you choose the right one for your use case.</p>-<h2>Speed Comparison</h2>-<p>Tested on 184 long-form audio files (192 minutes total). Higher RTF = faster.</p>+<h2>How to Compare Performance</h2>+<p>Throughput and latency depend on the GPU or CPU, precision, audio duration, batch size, concurrency, and runtime. Treat single-machine multipliers as workload-specific; benchmark RTF, p50/p95 latency, peak memory, and recognition quality on the target deployment.</p><table>-<thead><tr><th>Model</th><th>GPU Speed</th><th>CPU Speed</th><th>vs Whisper-large-v3</th></tr></thead>+<thead><tr><th>Deployment Goal</th><th>FunASR Path</th><th>Whisper Path</th><th>Measure</th></tr></thead><tbody>-<tr><td>FunASR Fun-ASR-Nano (vLLM)</td><td class="win">340x realtime</td><td>—</td><td><span class="badge">26x faster</span></td></tr>-<tr><td>FunASR SenseVoice-Small</td><td class="win">170x realtime</td><td class="win">17x realtime</td><td><span class="badge">13x faster</span></td></tr>-<tr><td>FunASR Paraformer-Large</td><td class="win">120x realtime</td><td class="win">15x realtime</td><td><span class="badge">9x faster</span></td></tr>-<tr><td>Whisper-large-v3-turbo</td><td>46x realtime</td><td class="lose">❌ Too slow</td><td>3.4x faster</td></tr>-<tr><td>Whisper-large-v3</td><td class="lose">13x realtime</td><td class="lose">❌</td><td>baseline</td></tr>+<tr><td>High-throughput GPU</td><td>Fun-ASR-Nano + vLLM</td><td>Whisper + batched runtime</td><td>Concurrency, batch, memory, and long-audio RTF</td></tr>+<tr><td>Low-latency CPU/GPU</td><td>SenseVoiceSmall or Paraformer</td><td>Quantized Whisper runtime</td><td>Single-request p50/p95, threads, and peak memory</td></tr>+<tr><td>Edge device</td><td>llama.cpp/GGUF or native runtime</td><td>whisper.cpp</td><td>Realtime behavior, power, and model size on-device</td></tr></tbody></table>-<p><strong>Key takeaway:</strong> FunASR models run on CPU faster than Whisper runs on GPU.</p>-<h2>Feature Comparison</h2><table><thead><tr><th>Feature</th><th>FunASR</th><th>Whisper</th></tr></thead><tbody>-<tr><td>Languages</td><td class="win">50+ (SenseVoice) / 31 (MLT-Nano) / zh-en-ja + Chinese dialects/accents (Nano)</td><td>57</td></tr>+<tr><td>Languages</td><td class="win">SenseVoiceSmall: zh/yue/en/ja/ko; MLT-Nano: 31 languages; Nano: zh/en/ja plus Chinese dialects and accents</td><td>57</td></tr><tr><td>Speaker Diarization</td><td class="win">✅ Built-in (cam++)</td><td>❌ Needs pyannote</td></tr><tr><td>Emotion Detection</td><td class="win">✅ Happy/Sad/Angry/Neutral</td><td>❌</td></tr><tr><td>Audio Event Detection</td><td class="win">✅ Music, applause, laughter</td><td>❌</td></tr>@@ -102,7 +98,7 @@ footer a{color:#94a3b8}<tr><td>OpenAI-compatible API</td><td class="win">✅ funasr-server</td><td>Separate wrapper needed</td></tr><tr><td>VAD (Voice Activity)</td><td class="win">✅ Built-in</td><td>❌ External</td></tr><tr><td>Punctuation</td><td class="win">✅ Built-in</td><td>Partial</td></tr>-<tr><td>CPU Inference</td><td class="win">✅ 17x realtime</td><td>❌ Impractical</td></tr>+<tr><td>CPU Inference</td><td class="win">✅ SenseVoice / Paraformer / native runtimes</td><td>✅ Quantized runtimes available</td></tr><tr><td>Fine-tuning</td><td class="win">✅ DeepSpeed scripts</td><td>Community scripts</td></tr><tr><td>License</td><td>MIT</td><td>MIT</td></tr><tr><td>Cost</td><td>Free (self-hosted)</td><td>Free (self-hosted)</td></tr>@@ -145,7 +141,7 @@ result = model.generate(input="meeting.wav")<div class="cta"><h3>Ready to try FunASR?</h3><p>The FunASR open-source community keeps growing. Try it and join the project.</p>-<a href="https://github.com/modelscope/FunASR" class="btn">View on GitHub ★</a>+<a href="/go/github" class="btn">View on GitHub ★</a></div><h2>Migration Guide</h2>@@ -155,11 +151,11 @@ result = model.generate(input="meeting.wav")<table><thead><tr><th>Project</th><th>Best For</th><th>Link</th></tr></thead><tbody>-<tr><td>FunASR</td><td>Full-featured toolkit (all models)</td><td><a href="https://github.com/modelscope/FunASR">GitHub</a></td></tr>-<tr><td>Fun-ASR-Nano</td><td>Flagship LLM-ASR for zh/en/ja + Chinese dialects/accents</td><td><a href="https://github.com/QwenAudio/Fun-ASR">GitHub</a></td></tr>-<tr><td>Fun-ASR-MLT-Nano</td><td>Separate multilingual checkpoint for 31 languages</td><td><a href="https://github.com/QwenAudio/Fun-ASR">GitHub</a></td></tr>-<tr><td>SenseVoice</td><td>Ultra-fast ASR + emotion + events</td><td><a href="https://github.com/QwenAudio/SenseVoice">GitHub</a></td></tr>-<tr><td>FunClip</td><td>AI video clipping with ASR</td><td><a href="https://github.com/modelscope/FunClip">GitHub</a></td></tr>+<tr><td>FunASR</td><td>Full-featured toolkit (all models)</td><td><a href="/go/github">GitHub</a></td></tr>+<tr><td>Fun-ASR-Nano</td><td>Flagship LLM-ASR for zh/en/ja + Chinese dialects/accents</td><td><a href="/go/fun-asr">GitHub</a></td></tr>+<tr><td>Fun-ASR-MLT-Nano</td><td>Separate multilingual checkpoint for 31 languages</td><td><a href="/go/fun-asr">GitHub</a></td></tr>+<tr><td>SenseVoice</td><td>Ultra-fast ASR + emotion + events</td><td><a href="/go/sensevoice">GitHub</a></td></tr>+<tr><td>FunClip</td><td>AI video clipping with ASR</td><td><a href="/go/funclip">GitHub</a></td></tr></tbody></table>diff --git a/web-pages/product-site/legacy/index.html b/web-pages/product-site/legacy/index.htmlindex f107d6ffe..27b370776 100644--- a/web-pages/product-site/legacy/index.html+++ b/web-pages/product-site/legacy/index.html@@ -4,7 +4,7 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>FunASR - 工业级语音识别服务 | 语音理解一站式解决方案</title>-<meta name="description" content="FunASR 是工业级语音识别服务,提供语音识别、语音检测、标点恢复、说话人分离、情感检测、音频事件识别,统一 Python 接口,支持 50+ 语言,私有化部署。">+<meta name="description" content="FunASR 是工业级语音识别服务,提供语音识别、语音检测、标点恢复、说话人分离、情感检测、音频事件识别,统一 Python 接口,并提供覆盖五语种 SenseVoiceSmall 与 31 语种 MLT-Nano 的私有化部署路径。"><link rel="icon" href="logo.png"><meta property="og:title" content="FunASR - 工业级语音理解一站式解决方案"><meta property="og:description" content="语音识别、语音检测、标点恢复、说话人分离、情感检测、音频事件识别——统一 Python 接口,一行命令部署。">@@ -17,7 +17,7 @@<link rel="alternate" hreflang="x-default" href="https://www.funasr.com/en/"><link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet"><script type="application/ld+json">-{"@context":"https://schema.org","@type":"SoftwareApplication","name":"FunASR","applicationCategory":"DeveloperApplication","operatingSystem":"Linux, Windows, macOS","offers":{"@type":"Offer","price":"0","priceCurrency":"USD"},"description":"Industrial-grade speech recognition toolkit supporting 50+ languages","url":"https://www.funasr.com","downloadUrl":"https://pypi.org/project/funasr/","author":{"@type":"Organization","name":"FunAudioLLM, Tongyi Lab"}}+{"@context":"https://schema.org","@type":"SoftwareApplication","name":"FunASR","applicationCategory":"DeveloperApplication","operatingSystem":"Linux, Windows, macOS","offers":{"@type":"Offer","price":"0","priceCurrency":"USD"},"description":"Industrial-grade speech recognition toolkit with model-specific language coverage and self-hosted deployment","url":"https://www.funasr.com","downloadUrl":"https://pypi.org/project/funasr/","author":{"@type":"Organization","name":"FunAudioLLM, Tongyi Lab"}}</script><script src="/lang-redirect.js"></script><style>@@ -141,7 +141,7 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}</div><a href="/en/" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">EN</a>-<a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a>+<a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><!-- Hero -->@@ -153,14 +153,14 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}<a href="https://huggingface.co/FunAudioLLM/Fun-ASR-Nano-2512" target="_blank" rel="noopener" class="hero-flagship">🚀 旗舰 <strong>Fun-ASR-Nano</strong> · 中/英/日 + 中文方言/口音</a><a href="https://huggingface.co/FunAudioLLM/Fun-ASR-MLT-Nano-2512" target="_blank" rel="noopener" class="hero-flagship mlt">🌍 31 语种 <strong>Fun-ASR-MLT-Nano</strong> · 独立 checkpoint</a></div>-<p class="hero-sub">语音识别、语音检测、标点恢复、说话人分离、情感检测、音频事件识别——统一的 Python 接口,一次调用完成全部处理。50+ 语言覆盖,私有化部署,生产就绪。</p>+<p class="hero-sub">语音识别、语音检测、标点恢复、说话人分离、情感检测、音频事件识别——统一的 Python 接口,一次调用完成全部处理。模型覆盖范围明确,支持私有化部署和生产服务。</p><div class="hero-actions"><a href="#usage" class="btn-primary">如何使用</a><a href="#demo" class="btn-outline">观看演示</a></div><div class="hero-stats">-<div class="hero-stat"><strong>50+</strong><span>支持语言</span></div>-<div class="hero-stat"><strong>170x</strong><span>实时处理速度</span></div>+<div class="hero-stat"><strong>31</strong><span>MLT-Nano 语种</span></div>+<div class="hero-stat"><strong>5</strong><span>SenseVoice 语种</span></div><div class="hero-stat"><strong>1 API</strong><span>统一调用接口</span></div></div></div>@@ -189,7 +189,7 @@ print(result.text)<p>从原始音频到结构化输出的完整语音理解流水线,一次调用全部搞定</p></div><div class="capabilities-grid">-<div class="cap-card"><div class="cap-icon">🎙️</div><h3>语音识别</h3><p>端到端 ASR,支持 50+ 语言,中文含 7 大方言及 26 种地域口音,自动语种检测</p></div>+<div class="cap-card"><div class="cap-icon">🎙️</div><h3>语音识别</h3><p>端到端 ASR;SenseVoiceSmall 覆盖中、粤、英、日、韩,MLT-Nano 覆盖 31 语种,Nano 面向中文方言与口音</p></div><div class="cap-card"><div class="cap-icon">📍</div><h3>语音检测 (VAD)</h3><p>毫秒级语音活动检测,自适应静音阈值,精确切分语音段与静音段</p></div><div class="cap-card"><div class="cap-icon">✍️</div><h3>标点恢复</h3><p>自动添加标点符号和逆文本正则化,输出可直接阅读的规范文本</p></div><div class="cap-card"><div class="cap-icon">👥</div><h3>说话人分离</h3><p>自动识别"谁说了什么",为每句话标注说话人 ID,支持多人会议</p></div>@@ -299,7 +299,7 @@ print(result.text)<div class="more-grid"><a class="more-item" href="https://github.com/modelscope/FunASR/blob/main/docs/vllm_guide_zh_v2.md" target="_blank">🔥 热词增强 <span>→</span></a><a class="more-item" href="https://modelscope.github.io/FunASR/zh/tutorial.html" target="_blank">👥 说话人分离 <span>→</span></a>-<a class="more-item" href="https://github.com/QwenAudio/SenseVoice" target="_blank">😊 情感检测 <span>→</span></a>+<a class="more-item" href="/go/sensevoice" target="_blank">😊 情感检测 <span>→</span></a><a class="more-item" href="https://modelscope.github.io/FunASR/zh/tutorial.html" target="_blank">🌍 多语言识别 <span>→</span></a><a class="more-item" href="https://modelscope.github.io/FunASR/zh/training.html" target="_blank">🎯 模型微调 <span>→</span></a><a class="more-item" href="https://github.com/modelscope/FunASR/blob/main/docs/vllm_guide_zh_v2.md" target="_blank">⚡ vLLM 加速 <span>→</span></a>@@ -350,13 +350,13 @@ print(result.text)<a href="https://modelscope.github.io/FunASR/api.html">API 参考</a></div><div class="footer-col"><h4>生态</h4>-<a href="https://github.com/QwenAudio/Fun-ASR">Fun-ASR-Nano</a>-<a href="https://github.com/QwenAudio/SenseVoice">SenseVoice</a>-<a href="https://github.com/modelscope/FunClip">FunClip</a>+<a href="/go/fun-asr">Fun-ASR-Nano</a>+<a href="/go/sensevoice">SenseVoice</a>+<a href="/go/funclip">FunClip</a><a href="https://github.com/QwenAudio/CosyVoice">CosyVoice</a></div><div class="footer-col"><h4>社区</h4>-<a href="https://github.com/modelscope/FunASR">GitHub</a>+<a href="/go/github">GitHub</a><a href="https://github.com/modelscope/FunASR/discussions">Discussions</a><a href="https://github.com/modelscope/FunASR/issues">问题反馈</a></div>diff --git a/web-pages/product-site/legacy/models.html b/web-pages/product-site/legacy/models.htmlindex bfa5a97a9..d3ef7a1e7 100644--- a/web-pages/product-site/legacy/models.html+++ b/web-pages/product-site/legacy/models.html@@ -4,7 +4,7 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>FunASR 模型 — SenseVoice、Paraformer、Fun-ASR-Nano、cam++</title>-<meta name="description" content="FunASR 模型选型指南:SenseVoice(170倍实时,50+语种)、Paraformer(中文高精度与可靠字级时间戳)、Fun-ASR-Nano(LLM 架构,vLLM 加速)、cam++(说话人分离)。">+<meta name="description" content="FunASR 模型选型指南:SenseVoice(中、粤、英、日、韩五语种)、Paraformer(中文高精度与可靠字级时间戳)、Fun-ASR-Nano(LLM 架构,vLLM 加速)、cam++(说话人分离)。"><link rel="icon" href="logo.png"><meta property="og:title" content="FunASR 模型 — 语音识别模型对比"><meta property="og:description" content="FunASR 模型选型:SenseVoice、Paraformer、Fun-ASR-Nano,按速度、精度、语种选择最适合的模型。">@@ -60,7 +60,7 @@ footer a{color:#60a5fa}</style></head><body>-<nav class="nav"><div class="container"><a class="nav-logo" href="./">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="quickstart.html">快速上手</a><a href="models.html">模型</a><a href="ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a><a href="en/models.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">EN</a><a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a></div></div></nav>+<nav class="nav"><div class="container"><a class="nav-logo" href="./">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="quickstart.html">快速上手</a><a href="models.html">模型</a><a href="ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a><a href="en/models.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">EN</a><a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></div></nav><article><div class="container">@@ -75,10 +75,10 @@ footer a{color:#60a5fa}<tr><th>模型</th><th>速度</th><th>语种</th><th>参数量</th><th>适用场景</th></tr></thead><tbody>-<tr><td><strong>Fun-ASR-Nano</strong> ⭐</td><td>vLLM 340x</td><td>中/英/日 + 中文方言/口音</td><td>800M</td><td>旗舰 · LLM-ASR · 难例最强</td></tr>+<tr><td><strong>Fun-ASR-Nano</strong> ⭐</td><td>vLLM;需按硬件实测</td><td>中/英/日 + 中文方言/口音</td><td>800M</td><td>旗舰 · LLM-ASR · 难例最强</td></tr><tr><td><strong>Fun-ASR-MLT-Nano</strong></td><td>vLLM</td><td>31</td><td>800M</td><td>独立多语 checkpoint</td></tr>-<tr><td><strong>SenseVoice Small</strong></td><td>170 倍实时</td><td>50+</td><td>234M</td><td>多语种、情感检测</td></tr>-<tr><td><strong>Paraformer-zh</strong></td><td>120 倍实时</td><td>中文、粤语</td><td>220M</td><td>中文最高精度</td></tr>+<tr><td><strong>SenseVoice Small</strong></td><td>非自回归;需按硬件实测</td><td>中、粤、英、日、韩</td><td>234M</td><td>五语种、情感检测</td></tr>+<tr><td><strong>Paraformer-zh</strong></td><td>非自回归;需按硬件实测</td><td>中文、粤语</td><td>220M</td><td>中文最高精度</td></tr><tr><td><strong>cam++</strong></td><td>实时</td><td>通用</td><td>7.2M</td><td>说话人分离与确认</td></tr></tbody></table>@@ -123,10 +123,10 @@ result = model.generate(input="audio.wav")</pre><div class="model-card"><div class="model-name">SenseVoice Small</div><div class="model-meta">234M 参数 · 非自回归 · <a href="/go/sensevoice">GitHub</a> · <a href="https://huggingface.co/FunAudioLLM/SenseVoiceSmall">HuggingFace</a></div>-<div class="model-desc">极速语音识别,内置情感和音频事件检测。支持 50+ 语种,包括中文、英文、日文、韩文、法文、德文等。非自回归架构在 GPU 上实现 170 倍实时速度。</div>+<div class="model-desc">极速语音识别,内置情感和音频事件检测。支持普通话、粤语、英语、日语和韩语。非自回归架构适合低延迟部署,实际吞吐取决于硬件、音频长度和 batch 配置。</div><div class="model-features">-<span class="badge badge-green">170 倍实时</span>-<span class="badge badge-blue">50+ 语种</span>+<span class="badge badge-green">非自回归</span>+<span class="badge badge-blue">五语种</span><span class="badge badge-orange">情感检测</span><span class="badge badge-blue">音频事件</span><span class="badge badge-green">支持 CPU</span>@@ -144,7 +144,7 @@ print(result[0]["text"])</pre><div class="model-meta">220M 参数 · 非自回归 · <a href="https://huggingface.co/funasr/paraformer-zh">HuggingFace</a></div><div class="model-desc">中文语音识别最高精度模型。基于 CTC 引导注意力的非自回归架构,使用 60,000+ 小时普通话语音训练。内置标点恢复和时间戳预测。</div><div class="model-features">-<span class="badge badge-green">120 倍实时</span>+<span class="badge badge-green">非自回归;需按硬件实测</span><span class="badge badge-blue">中文 + 粤语</span><span class="badge badge-green">最高精度</span><span class="badge badge-blue">时间戳</span>@@ -230,6 +230,6 @@ curl http://localhost:8000/v1/audio/transcriptions \</div></article>-<footer><div class="container">FunASR — 工业级语音识别工具包 — <a href="https://github.com/modelscope/FunASR">GitHub</a> · <a href="./">首页</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer>+<footer><div class="container">FunASR — 工业级语音识别工具包 — <a href="/go/github">GitHub</a> · <a href="./">首页</a> · <a href="https://www.funasr.com">funasr.com</a></div></footer></body></html>diff --git a/web-pages/product-site/legacy/quickstart.html b/web-pages/product-site/legacy/quickstart.htmlindex c68f39124..cced8180c 100644--- a/web-pages/product-site/legacy/quickstart.html+++ b/web-pages/product-site/legacy/quickstart.html@@ -141,7 +141,7 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}<a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="quickstart.html">快速上手</a><a href="models.html">模型</a><a href="ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="/en/quickstart.html" style="font-size:0.85rem;font-weight:500;color:var(--text-soft);text-decoration:none">EN</a>-<a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a>+<a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><!-- Hero -->@@ -178,7 +178,7 @@ footer{background:#0f172a;color:#94a3b8;padding:48px 0 32px}<section class="alt" id="file-transcription"><div class="container"><div class="guide-section"><h2><div class="step-number">1</div>离线文件转写</h2>-<p class="subtitle">上传音频文件,获得完整转录文本。支持 50+ 语言,自动包含情感检测和音频事件识别。</p>+<p class="subtitle">上传音频文件,获得完整转录文本。SenseVoiceSmall 支持中、粤、英、日、韩五种语言,并返回情感和音频事件标签。</p><div class="code-block"><span class="k">from</span> funasr <span class="k">import</span> AutoModel<br>@@ -254,7 +254,7 @@ chunk_stride = chunk_size[<span class="n">1</span>] * <span class="n">960</span><div class="tip-block"><strong>WebSocket 实时服务</strong>-如果需要浏览器端实时识别,FunASR 提供完整的 WebSocket 服务方案。参见 <a href="https://github.com/modelscope/FunASR" target="_blank">GitHub 仓库</a> 中的 <code>examples/</code> 目录获取 WebSocket 服务端和浏览器客户端代码。+如果需要浏览器端实时识别,FunASR 提供完整的 WebSocket 服务方案。参见 <a href="/go/github" target="_blank">GitHub 仓库</a> 中的 <code>examples/</code> 目录获取 WebSocket 服务端和浏览器客户端代码。</div></div></div></section>@@ -328,7 +328,7 @@ MCP (Model Context Protocol) 是 AI 助手调用外部工具的标准协议。<p>深入了解 FunASR 的更多能力</p></div><div class="next-grid">-<a class="next-card" href="https://github.com/modelscope/FunASR" target="_blank">+<a class="next-card" href="/go/github" target="_blank"><div class="next-icon">⭐</div><div><h3>GitHub 仓库</h3><p>源代码、示例、Issue 反馈</p></div></a>@@ -361,13 +361,13 @@ MCP (Model Context Protocol) 是 AI 助手调用外部工具的标准协议。<a href="https://modelscope.github.io/FunASR/api.html">API 参考</a></div><div class="footer-col"><h4>生态</h4>-<a href="https://github.com/QwenAudio/Fun-ASR">Fun-ASR-Nano</a>-<a href="https://github.com/QwenAudio/SenseVoice">SenseVoice</a>-<a href="https://github.com/modelscope/FunClip">FunClip</a>+<a href="/go/fun-asr">Fun-ASR-Nano</a>+<a href="/go/sensevoice">SenseVoice</a>+<a href="/go/funclip">FunClip</a><a href="https://github.com/QwenAudio/CosyVoice">CosyVoice</a></div><div class="footer-col"><h4>社区</h4>-<a href="https://github.com/modelscope/FunASR">GitHub</a>+<a href="/go/github">GitHub</a><a href="https://github.com/modelscope/FunASR/discussions">Discussions</a><a href="https://github.com/modelscope/FunASR/issues">问题反馈</a></div>diff --git a/web-pages/product-site/legacy/vs-whisper.html b/web-pages/product-site/legacy/vs-whisper.htmlindex 28424a828..c1cc9e553 100644--- a/web-pages/product-site/legacy/vs-whisper.html+++ b/web-pages/product-site/legacy/vs-whisper.html@@ -4,10 +4,10 @@<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>FunASR vs Whisper 对比 — 开源语音识别哪个好?</title>-<meta name="description" content="FunASR 与 Whisper 详细对比:FunASR 速度快 13 倍,内置说话人分离、情感检测、流式识别,支持 CPU 部署。">+<meta name="description" content="FunASR 与 Whisper 详细对比:模型覆盖、说话人分离、情感检测、流式识别、CPU/GPU 部署与可复现实测方法。"><link rel="icon" href="logo.png"><meta property="og:title" content="FunASR vs Whisper — 开源语音识别对比">-<meta property="og:description" content="FunASR 比 Whisper 快 13 倍,内置说话人分离和情感检测,支持 CPU 推理和实时流式。">+<meta property="og:description" content="按模型覆盖、说话人能力、情感检测、流式识别和部署条件对比 FunASR 与 Whisper。"><meta property="og:type" content="article"><meta property="og:url" content="https://www.funasr.com/vs-whisper.html"><meta property="og:image" content="https://www.funasr.com/logo.png">@@ -43,7 +43,7 @@ pre{background:#1e293b;color:#e2e8f0;padding:18px;border-radius:10px;overflow-x:.cta a.btn{display:inline-block;padding:10px 24px;background:var(--primary);color:white;border-radius:8px;font-weight:600;text-decoration:none}footer{background:#0f172a;color:#94a3b8;padding:28px 0;text-align:center;font-size:0.8rem}footer a{color:#94a3b8}-@media(max-width:600px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.7rem}table{font-size:0.8rem}th,td{padding:8px}}+@media(max-width:600px){.nav-links{display:none}.nav .container{gap:16px}.nav-logo{margin-right:auto}h1{font-size:1.7rem}table{font-size:0.8rem;table-layout:fixed}th,td{padding:8px;overflow-wrap:anywhere}}</style></head><body>@@ -51,7 +51,7 @@ footer a{color:#94a3b8}<a href="/" class="nav-logo">Fun<span>ASR</span></a><div class="nav-links"><a href="/#capabilities">核心能力</a><a href="/#usage">如何使用</a><a href="/#benchmark">性能</a><a href="/#demo">演示</a><a href="https://modelscope.github.io/FunASR/zh/" target="_blank">文档</a><a href="quickstart.html">快速上手</a><a href="models.html">模型</a><a href="ecosystem.html">生态</a><a href="/blog/">技术博客</a><a href="vs-whisper.html">对比 Whisper</a><a href="/donors.html">功德榜</a></div><a href="/en/vs-whisper.html" style="font-size:0.83rem;font-weight:500;color:var(--text-soft)">EN</a>-<a href="https://github.com/modelscope/FunASR" target="_blank" class="nav-btn">GitHub ★</a>+<a href="/go/github" target="_blank" class="nav-btn">GitHub ★</a></div></nav><article><div class="container">@@ -59,22 +59,18 @@ footer a{color:#94a3b8}<h1>FunASR vs Whisper:开源语音识别详细对比</h1><p>FunASR 和 OpenAI Whisper 都是开源语音识别工具。本文从速度、功能、部署等维度进行全面对比。</p>-<h2>速度对比</h2>-<p>测试条件:184 条长音频(共 192 分钟)。RTF 越高越快。</p>+<h2>性能比较方法</h2>+<p>吞吐和延迟会随 GPU/CPU 型号、精度、音频长度、batch、并发及运行时变化。不要把单一硬件上的倍速当成通用结论;请在目标环境中分别记录 RTF、p50/p95 延迟、峰值显存或内存,以及识别质量。</p><table>-<thead><tr><th>模型</th><th>GPU 速度</th><th>CPU 速度</th><th>vs Whisper-large-v3</th></tr></thead>+<thead><tr><th>部署目标</th><th>FunASR 路径</th><th>Whisper 路径</th><th>建议实测</th></tr></thead><tbody>-<tr><td>FunASR Fun-ASR-Nano (vLLM)</td><td class="win">340x 实时</td><td>—</td><td class="win">快 26 倍</td></tr>-<tr><td>FunASR SenseVoice-Small</td><td class="win">170x 实时</td><td class="win">17x 实时</td><td class="win">快 13 倍</td></tr>-<tr><td>FunASR Paraformer-Large</td><td class="win">120x 实时</td><td class="win">15x 实时</td><td class="win">快 9 倍</td></tr>-<tr><td>Whisper-large-v3-turbo</td><td>46x 实时</td><td>❌ 太慢</td><td>快 3.4 倍</td></tr>-<tr><td>Whisper-large-v3</td><td>13x 实时</td><td>❌</td><td>基准</td></tr>+<tr><td>高吞吐 GPU</td><td>Fun-ASR-Nano + vLLM</td><td>Whisper + 批处理运行时</td><td>并发、batch、显存和长音频 RTF</td></tr>+<tr><td>低延迟 CPU/GPU</td><td>SenseVoiceSmall 或 Paraformer</td><td>Whisper 量化实现</td><td>单请求 p50/p95、线程数和峰值内存</td></tr>+<tr><td>边缘设备</td><td>llama.cpp/GGUF 或原生 runtime</td><td>whisper.cpp</td><td>目标设备实时性、功耗和模型体积</td></tr></tbody></table>-<p><strong>结论:FunASR 在 CPU 上比 Whisper 在 GPU 上还快。</strong></p>-<h2>功能对比</h2><table><thead><tr><th>功能</th><th>FunASR</th><th>Whisper</th></tr></thead>@@ -85,11 +81,11 @@ footer a{color:#94a3b8}<tr><td>热词</td><td class="win">✅ 自定义词表增强</td><td>❌</td></tr><tr><td>中文方言</td><td class="win">7 大方言 + 26 地方口音</td><td>有限</td></tr><tr><td>OpenAI 兼容 API</td><td class="win">✅ funasr-server</td><td>需额外封装</td></tr>-<tr><td>CPU 推理</td><td class="win">✅ 17x 实时</td><td>❌ 不可用</td></tr>+<tr><td>CPU 推理</td><td class="win">✅ SenseVoice / Paraformer / 原生 runtime</td><td>✅ 量化实现可用</td></tr><tr><td>VAD</td><td class="win">✅ 内置</td><td>❌ 外部</td></tr><tr><td>标点恢复</td><td class="win">✅ 内置</td><td>部分</td></tr><tr><td>微调训练</td><td class="win">✅ DeepSpeed</td><td>社区脚本</td></tr>-<tr><td>语言数量</td><td>50+(SenseVoice)/ 31(MLT-Nano)/ 中英日 + 中文方言/口音(Nano)</td><td>57</td></tr>+<tr><td>语言数量</td><td>SenseVoiceSmall:中/粤/英/日/韩;MLT-Nano:31 语种;Nano:中/英/日及中文方言/口音</td><td>57</td></tr><tr><td>开源协议</td><td>MIT</td><td>MIT</td></tr></tbody></table>@@ -111,18 +107,18 @@ result = model.generate(input="meeting.wav")<div class="cta"><h3>开始使用 FunASR</h3><p>FunASR 开源社区持续增长,欢迎体验并参与共建。</p>-<a href="https://github.com/modelscope/FunASR" class="btn">GitHub ★ 立即体验</a>+<a href="/go/github" class="btn">GitHub ★ 立即体验</a></div><h2>生态项目</h2><table><thead><tr><th>项目</th><th>适用场景</th><th>链接</th></tr></thead><tbody>-<tr><td>FunASR</td><td>全功能工具包(所有模型)</td><td><a href="https://github.com/modelscope/FunASR">GitHub</a></td></tr>-<tr><td>Fun-ASR-Nano</td><td>旗舰 LLM-ASR,中英日 + 中文方言/口音</td><td><a href="https://github.com/QwenAudio/Fun-ASR">GitHub</a></td></tr>-<tr><td>Fun-ASR-MLT-Nano</td><td>独立多语 checkpoint,31 语言</td><td><a href="https://github.com/QwenAudio/Fun-ASR">GitHub</a></td></tr>-<tr><td>SenseVoice</td><td>超快 ASR + 情感 + 音频事件</td><td><a href="https://github.com/QwenAudio/SenseVoice">GitHub</a></td></tr>-<tr><td>FunClip</td><td>AI 智能视频剪辑</td><td><a href="https://github.com/modelscope/FunClip">GitHub</a></td></tr>+<tr><td>FunASR</td><td>全功能工具包(所有模型)</td><td><a href="/go/github">GitHub</a></td></tr>+<tr><td>Fun-ASR-Nano</td><td>旗舰 LLM-ASR,中英日 + 中文方言/口音</td><td><a href="/go/fun-asr">GitHub</a></td></tr>+<tr><td>Fun-ASR-MLT-Nano</td><td>独立多语 checkpoint,31 语言</td><td><a href="/go/fun-asr">GitHub</a></td></tr>+<tr><td>SenseVoice</td><td>超快 ASR + 情感 + 音频事件</td><td><a href="/go/sensevoice">GitHub</a></td></tr>+<tr><td>FunClip</td><td>AI 智能视频剪辑</td><td><a href="/go/funclip">GitHub</a></td></tr></tbody></table>diff --git a/web-pages/product-site/tests/browser/product-site.spec.ts b/web-pages/product-site/tests/browser/product-site.spec.tsindex ab15fae44..f0927a748 100644--- a/web-pages/product-site/tests/browser/product-site.spec.ts+++ b/web-pages/product-site/tests/browser/product-site.spec.ts@@ -222,3 +222,56 @@ test('llama.cpp blog heading clears fixed navigation on mobile', async ({ page }fullPage: true,});});++test('legacy comparison pages keep accurate claims and fit mobile', async ({ page }, testInfo) => {+ await page.setViewportSize({ width: 390, height: 844 });++ for (const route of ['/vs-whisper.html', '/en/vs-whisper.html']) {+ await page.goto(route);+ const audit = await page.evaluate(() => ({+ overflow: document.documentElement.scrollWidth - document.documentElement.clientWidth,+ text: document.body.innerText,+ trackedGitHub: document.querySelector('.nav-btn')?.getAttribute('href'),+ }));++ expect(audit.overflow).toBeLessThanOrEqual(1);+ expect(audit.text).not.toMatch(/50\+\s*(?:supported\s+|支持\s*)?(?:languages?|语言|语种)/i);+ expect(audit.text).not.toMatch(/(?:13|15|17|170)[x×]|(?:13|15|17|170)\s*倍/);+ expect(audit.trackedGitHub).toBe('/go/github');+ }++ await page.screenshot({+ path: testInfo.outputPath('legacy-comparison-mobile.png'),+ fullPage: true,+ });+});++test('SenseVoice guides keep mobile navigation clear of the article', async ({ page }, testInfo) => {+ await page.setViewportSize({ width: 390, height: 844 });++ for (const route of [+ '/blog/sensevoice-deployment-guide.html',+ '/en/blog/sensevoice-deployment-guide.html',+ ]) {+ await page.goto(route);+ const layout = await page.evaluate(() => {+ const navigation = document.querySelector<HTMLElement>('nav.nav');+ const heading = document.querySelector<HTMLElement>('h1');+ return {+ overflow: document.documentElement.scrollWidth - document.documentElement.clientWidth,+ navigationBottom: navigation?.getBoundingClientRect().bottom ?? 0,+ headingTop: heading?.getBoundingClientRect().top ?? 0,+ trackedSenseVoice: document.querySelector('.cta a')?.getAttribute('href'),+ };+ });++ expect(layout.overflow).toBeLessThanOrEqual(1);+ expect(layout.headingTop).toBeGreaterThanOrEqual(layout.navigationBottom + 16);+ expect(layout.trackedSenseVoice).toBe('/go/sensevoice');+ }++ await page.screenshot({+ path: testInfo.outputPath('sensevoice-guide-mobile.png'),+ fullPage: true,+ });+});diff --git a/web-pages/product-site/tests/test_legacy.py b/web-pages/product-site/tests/test_legacy.pyindex a39d444bf..525648a76 100644--- a/web-pages/product-site/tests/test_legacy.py+++ b/web-pages/product-site/tests/test_legacy.py@@ -2,6 +2,7 @@ from __future__ import annotationsimport hashlibimport json+import reimport sysfrom pathlib import Path@@ -95,6 +96,7 @@ def test_navigation_normalization_is_idempotent_and_preserves_article():assert once == twiceassert article in onceassert [link.get_text(strip=True) for link in soup.select('.nav-links a')][-1] == '功德榜'+ assert soup.select_one('.nav-btn')['href'] == '/go/github'assert soup.select_one('link[rel="canonical"]')['href'].endswith('/blog/example.html')assert not soup.select_one('link[href*="fonts.googleapis.com"]')assert not soup.select_one('script[src="/stats/tracker.js"]')@@ -178,3 +180,26 @@ def test_llama_cpp_blog_points_directly_to_current_runtime_release():assert expected_routes[language] in hrefsassert '2026-08-15' in soup.select_one('script[type="application/ld+json"]').stringassert '**' not in soup.select_one('article').get_text()+++def test_public_pages_do_not_overstate_sensevoice_language_or_speed_claims():+ language_claim = re.compile(+ r'50\+\s*(?:supported\s+|支持\s*)?(?:languages?|语言|语种)',+ re.IGNORECASE,+ )+ fixed_speed_claim = re.compile(+ r'\b(?:13|15|17|170)[x×]\b|(?:13|15|17|170)\s*倍'+ )++ violations = []+ for path in sorted(LEGACY.rglob('*.html')):+ source = path.read_text(encoding='utf-8')+ text = BeautifulSoup(source, 'html.parser').get_text(+ ' ',+ strip=True,+ )+ searchable = f'{source}\n{text}'+ if language_claim.search(searchable) or fixed_speed_claim.search(searchable):+ violations.append(path.relative_to(LEGACY).as_posix())++ assert violations == []