Voice-to-text dictation app with local (Nvidia Parakeet/Whisper) and cloud models (BYOK). Privacy-first and available cross-platform.
JavaScript aianthropiccross-platformgeminigroqlinuxmacosnvidiaopen-sourceopenaiparakeetspeech-to-texttranscribewhisperwindows
{
"parakeetModels": {
"parakeet-tdt-0.6b-v3": {
"name": "Parakeet TDT 0.6B",
"sherpaModelType": "nemo_transducer",
"description": "Fast multilingual ASR with auto language detection",
"size": "680MB",
"sizeMb": 680,
"language": "multilingual",
"supportedLanguages": [
"bg",
"hr",
"cs",
"da",
"nl",
"en",
"et",
"fi",
"fr",
"de",
"el",
"hu",
"it",
"lv",
"lt",
"mt",
"pl",
"pt",
"ro",
"sk",
"sl",
"es",
"sv",
"ru",
"uk"
],
"downloadUrl": "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8.tar.bz2",
"extractDir": "sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8",
"descriptionKey": "models.descriptions.parakeet.parakeet_tdt_0_6b_v3"
},
"parakeet-unified-en-0.6b": {
"name": "Parakeet Unified EN 0.6B",
"description": "English-only ASR with state-of-the-art accuracy (5.91% avg WER)",
"size": "631MB",
"sizeMb": 631,
"language": "en",
"supportedLanguages": ["en"],
"downloadUrl": "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-nemo-parakeet-unified-en-0.6b-int8-non-streaming.tar.bz2",
"extractDir": "sherpa-onnx-nemo-parakeet-unified-en-0.6b-int8-non-streaming",
"descriptionKey": "models.descriptions.parakeet.parakeet_unified_en_0_6b"
},
"nemotron-speech-streaming-en-0.6b": {
"name": "Nemotron Speech Streaming EN 0.6B",
"description": "English streaming ASR from NVIDIA Nemotron",
"size": "632MB",
"sizeMb": 632,
"language": "en",
"supportedLanguages": ["en"],
"runtime": "online",
"downloadUrl": "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-nemotron-speech-streaming-en-0.6b-560ms-int8-2026-04-25.tar.bz2",
"extractDir": "sherpa-onnx-nemotron-speech-streaming-en-0.6b-560ms-int8-2026-04-25",
"descriptionKey": "models.descriptions.parakeet.nemotron_speech_streaming_en_0_6b"
},
"nemotron-3.5-asr-streaming-0.6b": {
"name": "Nemotron 3.5 ASR Streaming 0.6B",
"description": "Multilingual streaming ASR from NVIDIA Nemotron with auto language detection (15 languages)",
"size": "650MB",
"sizeMb": 650,
"language": "multilingual",
"supportedLanguages": [
"en",
"es",
"fr",
"it",
"pt",
"nl",
"de",
"tr",
"ru",
"ar",
"hi",
"ja",
"ko",
"vi",
"uk"
],
"runtime": "online",
"downloadUrl": "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-nemotron-3.5-asr-streaming-0.6b-560ms-int8-2026-06-11.tar.bz2",
"extractDir": "sherpa-onnx-nemotron-3.5-asr-streaming-0.6b-560ms-int8-2026-06-11",
"descriptionKey": "models.descriptions.parakeet.nemotron_3_5_asr_streaming_0_6b"
},
"cohere-transcribe-03-2026": {
"name": "Cohere Transcribe 2B",
"description": "Most accurate local ASR (5.42% avg WER, 14 languages). No auto language detection",
"size": "2.7GB",
"sizeMb": 2700,
"expectedSizeBytes": 1699791751,
"language": "multilingual",
"supportedLanguages": [
"ar",
"de",
"el",
"en",
"es",
"fr",
"it",
"ja",
"ko",
"nl",
"pl",
"pt",
"vi",
"zh"
],
"modelType": "cohere-transcribe",
"downloadUrl": "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-cohere-transcribe-14-lang-int8-2026-04-01.tar.bz2",
"extractDir": "sherpa-onnx-cohere-transcribe-14-lang-int8-2026-04-01",
"descriptionKey": "models.descriptions.cohere.cohere_transcribe_03_2026"
},
"orukeet-v0.1.0": {
"name": "Orukeet",
"description": "Multilingual Parakeet fine-tune with frozen Gabor kernels",
"descriptionKey": "models.descriptions.parakeet.orukeet",
"size": "672MB",
"sizeMb": 671.6198,
"expectedSizeBytes": 486807585,
"language": "multilingual",
"supportedLanguages": [
"bg",
"hr",
"cs",
"da",
"nl",
"en",
"et",
"fi",
"fr",
"de",
"el",
"hu",
"it",
"lv",
"lt",
"mt",
"pl",
"pt",
"ro",
"sk",
"sl",
"es",
"sv",
"ru",
"uk"
],
"runtime": "offline",
"sherpaModelType": "nemo_transducer",
"recommended": true,
"extractDir": "sherpa-onnx-orukeet-v0.1.0-int8",
"downloadUrl": "https://huggingface.co/oruk/orukeet/resolve/55a984d46f68323301837194ce647c702f55facc/onnx/sherpa-onnx-orukeet-v0.1.0-int8.tar.bz2",
"manifestUrl": "https://huggingface.co/oruk/orukeet/resolve/55a984d46f68323301837194ce647c702f55facc/onnx/manifest.json",
"license": "CC-BY-SA-4.0",
"organization": {
"id": "oruk",
"name": "Oruk"
},
"modelCardUrl": "https://huggingface.co/oruk/orukeet"
}
},
"diarizationModels": {
"sherpa-onnx-pyannote-segmentation-3-0": {
"name": "Pyannote Segmentation 3.0",
"description": "Speech boundary detection for speaker diarization",
"size": "6.6 MB",
"sizeMb": 6.6,
"downloadUrl": "https://github.com/k2-fsa/sherpa-onnx/releases/download/speaker-segmentation-models/sherpa-onnx-pyannote-segmentation-3-0.tar.bz2",
"extractDir": "sherpa-onnx-pyannote-segmentation-3-0",
"descriptionKey": "models.descriptions.diarization.segmentation"
},
"3dspeaker-campplus-en": {
"name": "CAM++ Speaker Embedding (English)",
"description": "Speaker voice fingerprinting for identification",
"size": "28.2 MB",
"sizeMb": 28.2,
"downloadUrl": "https://github.com/k2-fsa/sherpa-onnx/releases/download/speaker-recongition-models/3dspeaker_speech_campplus_sv_en_voxceleb_16k.onnx",
"fileName": "3dspeaker_speech_campplus_sv_en_voxceleb_16k.onnx",
"descriptionKey": "models.descriptions.diarization.embedding_en"
}
},
"whisperModels": {
"tiny": {
"name": "Tiny",
"description": "Fastest, lower quality",
"size": "75MB",
"sizeMb": 75,
"expectedSizeBytes": 78000000,
"fileName": "ggml-tiny.bin",
"downloadUrl": "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-tiny.bin",
"descriptionKey": "models.descriptions.whisper.tiny"
},
"base": {
"name": "Base",
"description": "Good balance",
"size": "142MB",
"sizeMb": 142,
"expectedSizeBytes": 148000000,
"fileName": "ggml-base.bin",
"downloadUrl": "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin",
"recommended": true,
"descriptionKey": "models.descriptions.whisper.base"
},
"small": {
"name": "Small",
"description": "Better quality, slower",
"size": "466MB",
"sizeMb": 466,
"expectedSizeBytes": 488000000,
"fileName": "ggml-small.bin",
"downloadUrl": "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-small.bin",
"descriptionKey": "models.descriptions.whisper.small"
},
"medium": {
"name": "Medium",
"description": "High quality",
"size": "1.5GB",
"sizeMb": 1500,
"expectedSizeBytes": 1570000000,
"fileName": "ggml-medium.bin",
"downloadUrl": "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-medium.bin",
"descriptionKey": "models.descriptions.whisper.medium"
},
"large": {
"name": "Large",
"description": "Best quality, slowest",
"size": "3GB",
"sizeMb": 3000,
"expectedSizeBytes": 3140000000,
"fileName": "ggml-large-v3.bin",
"downloadUrl": "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-large-v3.bin",
"descriptionKey": "models.descriptions.whisper.large"
},
"turbo": {
"name": "Turbo",
"description": "Fast with good quality",
"size": "1.6GB",
"sizeMb": 1600,
"expectedSizeBytes": 1670000000,
"fileName": "ggml-large-v3-turbo.bin",
"downloadUrl": "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-large-v3-turbo.bin",
"descriptionKey": "models.descriptions.whisper.turbo"
}
},
"transcriptionProviders": [
{
"id": "openai",
"name": "OpenAI",
"baseUrl": "https://api.openai.com/v1",
"models": [
{
"id": "gpt-transcribe",
"name": "GPT Transcribe",
"description": "Newest, most accurate transcription",
"descriptionKey": "models.descriptions.transcription.openai_gpt_transcribe"
},
{
"id": "gpt-4o-mini-transcribe",
"name": "GPT-4o Mini Transcribe",
"description": "Fast and accurate transcription",
"descriptionKey": "models.descriptions.transcription.openai_gpt_4o_mini_transcribe",
"streaming": true
},
{
"id": "gpt-4o-transcribe",
"name": "GPT-4o Transcribe",
"description": "Most accurate transcription",
"descriptionKey": "models.descriptions.transcription.openai_gpt_4o_transcribe",
"streaming": true
},
{
"id": "whisper-1",
"name": "Whisper",
"description": "Original Whisper model",
"descriptionKey": "models.descriptions.transcription.openai_whisper_1"
}
]
},
{
"id": "groq",
"name": "Groq",
"baseUrl": "https://api.groq.com/openai/v1",
"models": [
{
"id": "whisper-large-v3",
"name": "Whisper Large v3",
"description": "High accuracy speech recognition",
"descriptionKey": "models.descriptions.transcription.groq_whisper_large_v3"
},
{
"id": "whisper-large-v3-turbo",
"name": "Whisper Large v3 Turbo",
"description": "216x real-time speed",
"descriptionKey": "models.descriptions.transcription.groq_whisper_large_v3_turbo"
}
]
},
{
"id": "xai",
"name": "xAI",
"baseUrl": "https://api.x.ai/v1",
"models": [
{
"id": "grok-stt",
"name": "Grok STT",
"description": "High-accuracy speech recognition by xAI",
"descriptionKey": "models.descriptions.transcription.xai_grok_stt"
}
]
},
{
"id": "mistral",
"name": "Mistral",
"baseUrl": "https://api.mistral.ai/v1",
"models": [
{
"id": "voxtral-mini-latest",
"name": "Voxtral Mini",
"description": "Fast multilingual transcription",
"descriptionKey": "models.descriptions.transcription.mistral_voxtral_mini_latest"
}
]
},
{
"id": "gemini",
"name": "Gemini",
"baseUrl": "https://generativelanguage.googleapis.com/v1beta",
"batchModel": "gemini-3.5-transcribe",
"models": [
{
"id": "gemini-3.5-transcribe",
"name": "Gemini 3.5 Transcribe",
"description": "Google's Gemini speech-to-text",
"descriptionKey": "models.descriptions.transcription.gemini_3_5_transcribe"
},
{
"id": "gemini-3.5-transcribe-live",
"name": "Gemini 3.5 Transcribe Live",
"description": "Realtime streaming speech-to-text",
"descriptionKey": "models.descriptions.transcription.gemini_3_5_transcribe_live",
"streaming": true
}
]
},
{
"id": "corti",
"name": "Corti",
"baseUrl": "https://api.us.corti.app/v2",
"models": [
{
"id": "corti-transcribe",
"name": "Corti Transcribe",
"description": "Clinical-grade medical transcription",
"descriptionKey": "models.descriptions.transcription.corti_transcribe",
"streaming": true
}
]
},
{
"id": "tinfoil",
"name": "Tinfoil",
"baseUrl": "https://inference.tinfoil.sh/v1",
"batchModel": "voxtral-small-24b",
"models": [
{
"id": "voxtral-mini-4b-realtime",
"name": "Voxtral",
"description": "Verifiably private transcription",
"descriptionKey": "models.descriptions.transcription.tinfoil_voxtral",
"streaming": true
}
]
},
{
"id": "deepgram",
"name": "Deepgram",
"baseUrl": "https://api.deepgram.com/v1",
"models": [
{
"id": "nova-3",
"name": "Nova-3",
"description": "Deepgram's most accurate realtime model",
"descriptionKey": "models.descriptions.transcription.deepgram_nova_3",
"streaming": true
}
]
},
{
"id": "assemblyai",
"name": "AssemblyAI",
"baseUrl": "https://api.assemblyai.com/v2",
"models": [
{
"id": "universal-3-5-pro",
"name": "Universal-3.5 Pro",
"description": "Highest accuracy realtime, 18 languages",
"descriptionKey": "models.descriptions.transcription.assemblyai_universal_3_5_pro",
"streaming": true
},
{
"id": "universal-streaming-english",
"name": "Universal Streaming (English)",
"description": "Immutable realtime transcripts, English",
"descriptionKey": "models.descriptions.transcription.assemblyai_universal_streaming_english",
"streaming": true
},
{
"id": "universal-streaming-multilingual",
"name": "Universal Streaming (Multilingual)",
"description": "Immutable realtime transcripts, multilingual",
"descriptionKey": "models.descriptions.transcription.assemblyai_universal_streaming_multilingual",
"streaming": true
}
]
}
],
"cloudProviders": [
{
"id": "openai",
"name": "OpenAI",
"models": [
{
"id": "gpt-6-astra",
"name": "GPT-6 Astra",
"description": "OpenAI's most capable model for coding, computer use, and research, 1M context",
"descriptionKey": "models.descriptions.cloud.openai_gpt_6_astra",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["low", "medium", "high", "xhigh", "max"]
}
},
{
"id": "gpt-5.6-sol",
"name": "GPT-5.6 Sol",
"description": "Flagship GPT-5.6 for complex coding, research, and agentic work, 1M context",
"descriptionKey": "models.descriptions.cloud.openai_gpt_5_6_sol",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["none", "low", "medium", "high", "xhigh", "max"],
"defaultValue": "medium"
}
},
{
"id": "gpt-5.6-terra",
"name": "GPT-5.6 Terra",
"description": "Balanced capability, speed, and cost for everyday work",
"descriptionKey": "models.descriptions.cloud.openai_gpt_5_6_terra",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["none", "low", "medium", "high", "xhigh", "max"],
"defaultValue": "medium"
}
},
{
"id": "gpt-5.6-luna",
"name": "GPT-5.6 Luna",
"description": "Fastest and lowest-cost model in the GPT-5.6 family",
"descriptionKey": "models.descriptions.cloud.openai_gpt_5_6_luna",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["none", "low", "medium", "high", "xhigh", "max"],
"defaultValue": "medium"
}
},
{
"id": "gpt-5.5",
"name": "GPT-5.5",
"description": "Frontier model for complex reasoning, 1M context",
"descriptionKey": "models.descriptions.cloud.openai_gpt_5_5",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["none", "low", "medium", "high", "xhigh"],
"defaultValue": "medium"
}
},
{
"id": "gpt-5.2",
"name": "GPT-5.2",
"description": "Strong reasoning model",
"descriptionKey": "models.descriptions.cloud.openai_gpt_5_2",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["none", "low", "medium", "high", "xhigh"],
"defaultValue": "none"
}
},
{
"id": "gpt-5-mini",
"name": "GPT-5 Mini",
"description": "Fast and cost-efficient",
"descriptionKey": "models.descriptions.cloud.openai_gpt_5_mini",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["minimal", "low", "medium", "high"],
"defaultValue": "medium"
}
},
{
"id": "gpt-5-nano",
"name": "GPT-5 Nano",
"description": "Ultra-fast, low latency",
"descriptionKey": "models.descriptions.cloud.openai_gpt_5_nano",
"tokenParam": "max_completion_tokens",
"supportsTemperature": false,
"supportsVision": true,
"reasoningEffort": {
"options": ["minimal", "low", "medium", "high"],
"defaultValue": "medium"
}
},
{
"id": "gpt-4.1",
"name": "GPT-4.1",
"description": "Strong baseline, 1M context",
"descriptionKey": "models.descriptions.cloud.openai_gpt_4_1",
"tokenParam": "max_completion_tokens",
"supportsTemperature": true,
"supportsVision": true
},
{
"id": "gpt-4.1-mini",
"name": "GPT-4.1 Mini",
"description": "Smaller GPT-4.1 model",
"descriptionKey": "models.descriptions.cloud.openai_gpt_4_1_mini",
"tokenParam": "max_completion_tokens",
"supportsTemperature": true,
"supportsVision": true
},
{
"id": "gpt-4.1-nano",
"name": "GPT-4.1 Nano",
"description": "Lowest latency GPT-4.1",
"descriptionKey": "models.descriptions.cloud.openai_gpt_4_1_nano",
"tokenParam": "max_completion_tokens",
"supportsTemperature": true,
"supportsVision": true
}
]
},
{
"id": "anthropic",
"name": "Anthropic",
"models": [
{
"id": "claude-fable-5-1",
"name": "Claude Fable 5.1",
"description": "Anthropic's most capable model, Mythos-class, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_fable_5_1",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-fable-5",
"name": "Claude Fable 5",
"description": "Previous Mythos-class flagship, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_fable_5",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-opus-5-5",
"name": "Claude Opus 5.5",
"description": "Latest Opus model for complex reasoning, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_opus_5_5",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-sonnet-5-5",
"name": "Claude Sonnet 5.5",
"description": "Latest Sonnet: fast and capable for everyday work, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_sonnet_5_5",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-haiku-5-5",
"name": "Claude Haiku 5.5",
"description": "Fastest, lowest-cost Claude model, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_haiku_5_5",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-sonnet-5",
"name": "Claude Sonnet 5",
"description": "Fast, capable agentic model at lower cost",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_sonnet_5",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-sonnet-4-6",
"name": "Claude Sonnet 4.6",
"description": "Balanced performance",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_sonnet_4_6",
"supportsTemperature": true,
"supportsVision": true
},
{
"id": "claude-haiku-4-5",
"name": "Claude Haiku 4.5",
"description": "Fast with near-frontier intelligence",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_haiku_4_5",
"supportsTemperature": true,
"supportsVision": true
},
{
"id": "claude-opus-5",
"name": "Claude Opus 5",
"description": "Previous Opus flagship, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_opus_5",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-opus-4-8",
"name": "Claude Opus 4.8",
"description": "Powerful Opus model tuned for honesty and reliability, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_opus_4_8",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-opus-4-7",
"name": "Claude Opus 4.7",
"description": "Powerful Opus model, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_opus_4_7",
"supportsTemperature": false,
"supportsVision": true
},
{
"id": "claude-opus-4-6",
"name": "Claude Opus 4.6",
"description": "Previous Opus generation, 1M context",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_opus_4_6",
"supportsTemperature": true,
"supportsVision": true
},
{
"id": "claude-sonnet-4-5",
"name": "Claude Sonnet 4.5",
"description": "Previous Sonnet generation",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_sonnet_4_5",
"supportsTemperature": true,
"supportsVision": true
},
{
"id": "claude-opus-4-5",
"name": "Claude Opus 4.5",
"description": "Earlier Opus model",
"descriptionKey": "models.descriptions.cloud.anthropic_claude_opus_4_5",
"supportsTemperature": true,
"supportsVision": true
}
]
},
{
"id": "gemini",
"name": "Google Gemini",
"models": [
{
"id": "gemini-3.5-flash",
"name": "Gemini 3.5 Flash",
"description": "Latest fast, high-capability Gemini model",
"descriptionKey": "models.descriptions.cloud.gemini_gemini_3_5_flash",
"supportsThinking": true,
"supportsVision": true,
"reasoningEffort": {
"options": ["minimal", "low", "medium", "high"],
"defaultValue": "medium"
}
},
{
"id": "gemini-3.5-flash-lite",
"name": "Gemini 3.5 Flash Lite",
"description": "Fastest, most cost-effective 3.5 model",
"descriptionKey": "models.descriptions.cloud.gemini_gemini_3_5_flash_lite",
"supportsThinking": true,
"supportsVision": true,
"reasoningEffort": {
"options": ["minimal", "low", "medium", "high"],
"defaultValue": "minimal"
}
},
{
"id": "gemini-3.1-pro-preview",
"name": "Gemini 3.1 Pro",
"description": "Next-gen flagship model for complex reasoning",
"descriptionKey": "models.descriptions.cloud.gemini_gemini_3_1_pro_preview",
"supportsThinking": true,
"supportsVision": true,
"reasoningEffort": {
"options": ["low", "medium", "high"],
"defaultValue": "high"
}
},
{
"id": "gemini-3.1-flash-lite",
"name": "Gemini 3.1 Flash Lite",
"description": "Frontier-class performance at low cost",
"descriptionKey": "models.descriptions.cloud.gemini_gemini_3_1_flash_lite",
"supportsVision": true,
"reasoningEffort": {
"options": ["minimal", "low", "medium", "high"]
}
},
{
"id": "gemini-3-flash-preview",
"name": "Gemini 3 Flash",
"description": "Ultra-fast, high-capability next-gen model",
"descriptionKey": "models.descriptions.cloud.gemini_gemini_3_flash_preview",
"supportsThinking": true,
"supportsVision": true,
"reasoningEffort": {
"options": ["minimal", "low", "medium", "high"],
"defaultValue": "high"
}
},
{
"id": "gemini-2.5-flash-lite",
"name": "Gemini 2.5 Flash Lite",
"description": "Lowest latency and cost",
"descriptionKey": "models.descriptions.cloud.gemini_gemini_2_5_flash_lite",
"supportsVision": true
},
{
"id": "gemma-4-31b-it",
"name": "Gemma 4 31B",
"description": "Google's largest Gemma 4, 31B dense, 256K context",
"descriptionKey": "models.descriptions.cloud.gemini_gemma_4_31b_it"
},
{
"id": "gemma-4-26b-a4b-it",
"name": "Gemma 4 26B MoE",
"description": "Google's Gemma 4 MoE, 4B active params, 256K context",
"descriptionKey": "models.descriptions.cloud.gemini_gemma_4_26b_a4b_it"
}
]
},
{
"id": "groq",
"name": "Groq",
"models": [
{
"id": "openai/gpt-oss-120b",
"name": "GPT-OSS 120B",
"description": "OpenAI's open-source flagship, 500 T/sec",
"supportsThinking": true,
"descriptionKey": "models.descriptions.cloud.groq_openai_gpt_oss_120b",
"reasoningEffort": {
"options": ["low", "medium", "high"]
}
},
{
"id": "qwen/qwen3.8-27b",
"name": "Qwen3.8 27B",
"description": "Fast Qwen model, 450 T/sec, 131K context",
"disableThinking": true,
"descriptionKey": "models.descriptions.cloud.groq_qwen_qwen3_8_27b"
},
{
"id": "openai/gpt-oss-20b",
"name": "GPT-OSS 20B",
"description": "Fast open-source model, 1000 T/sec",
"supportsThinking": true,
"descriptionKey": "models.descriptions.cloud.groq_openai_gpt_oss_20b",
"reasoningEffort": {
"options": ["low", "medium", "high"]
}
}
]
},
{
"id": "tinfoil",
"name": "Tinfoil",
"defaultModel": "glm-5-3",
"models": [
{
"id": "glm-5-3",
"name": "GLM-5.3",
"description": "Private agentic coding model in secure enclaves, 1M context",
"tokenParam": "max_tokens",
"supportsTemperature": true,
"descriptionKey": "models.descriptions.cloud.tinfoil_glm_5_3",
"supportsThinking": true
},
{
"id": "gpt-oss-120b",
"name": "GPT-OSS 120B",
"description": "Private open-weight reasoning model with tool use, 131K context",
"tokenParam": "max_tokens",
"supportsTemperature": true,
"descriptionKey": "models.descriptions.cloud.tinfoil_gpt_oss_120b",
"supportsThinking": true
},
{
"id": "gemma4-31b",
"name": "Gemma 4 31B",
"description": "Private multilingual reasoning model with tool calling, 256K context",
"tokenParam": "max_tokens",
"supportsTemperature": true,
"descriptionKey": "models.descriptions.cloud.tinfoil_gemma4_31b",
"supportsThinking": true
}
]
},
{
"id": "corti",
"name": "Corti",
"models": [
{
"id": "corti-s1-instant",
"name": "Corti S1 Instant",
"description": "Fast healthcare model, no reasoning trace",
"descriptionKey": "models.descriptions.cloud.corti_corti_s1_instant",
"tokenParam": "max_tokens",
"supportsTemperature": true
},
{
"id": "corti-s1",
"name": "Corti S1",
"description": "Most capable clinical model with reasoning",
"descriptionKey": "models.descriptions.cloud.corti_corti_s1",
"tokenParam": "max_tokens",
"supportsTemperature": true
},
{
"id": "corti-s1-mini-instant",
"name": "Corti S1 Mini Instant",
"description": "Compact healthcare model, fastest and lowest cost",
"descriptionKey": "models.descriptions.cloud.corti_corti_s1_mini_instant",
"tokenParam": "max_tokens",
"supportsTemperature": true
},
{
"id": "corti-s1-mini",
"name": "Corti S1 Mini",
"description": "Compact clinical model with reasoning at half the price",
"descriptionKey": "models.descriptions.cloud.corti_corti_s1_mini",
"tokenParam": "max_tokens",
"supportsTemperature": true
}
]
}
],
"enterpriseProviders": [
{
"id": "bedrock",
"name": "AWS Bedrock",
"allowCustomModelId": true,
"models": [
{
"id": "us.anthropic.claude-fable-5",
"name": "Claude Fable 5",
"description": "Most capable Claude model, Mythos-class. Best for the hardest Agent Mode tasks.",
"descriptionKey": "models.descriptions.enterprise.bedrock_claude_fable_5"
},
{
"id": "us.anthropic.claude-haiku-4-5-20251001-v1:0",
"name": "Claude Haiku 4.5",
"description": "Fast and cheap. Great default for text cleanup.",
"descriptionKey": "models.descriptions.enterprise.bedrock_claude_haiku_4_5"
},
{
"id": "us.anthropic.claude-sonnet-5",
"name": "Claude Sonnet 5",
"description": "Balanced workhorse. Best default for Agent Mode.",
"descriptionKey": "models.descriptions.enterprise.bedrock_claude_sonnet_5"
},
{
"id": "us.anthropic.claude-opus-4-8",
"name": "Claude Opus 4.8",
"description": "Max intelligence. Best for complex Agent Mode tasks.",
"descriptionKey": "models.descriptions.enterprise.bedrock_claude_opus_4_8"
},
{
"id": "openai.gpt-oss-120b-1:0",
"name": "GPT-OSS 120B",
"description": "Open-weight OpenAI model. Strong reasoning at low cost.",
"descriptionKey": "models.descriptions.enterprise.bedrock_gpt_oss_120b"
},
{
"id": "deepseek.v3.2",
"name": "DeepSeek V3.2",
"description": "Open-weight frontier model. Great value for Agent Mode.",
"descriptionKey": "models.descriptions.enterprise.bedrock_deepseek_v3_2"
},
{
"id": "qwen.qwen3-next-80b-a3b",
"name": "Qwen3 Next 80B",
"description": "Sparse open-weight model. Fast and multilingual.",
"descriptionKey": "models.descriptions.enterprise.bedrock_qwen3_next_80b"
}
]
},
{
"id": "azure",
"name": "Azure OpenAI",
"allowCustomModelId": true,
"models": []
},
{
"id": "vertex",
"name": "GCP Vertex AI",
"allowCustomModelId": true,
"models": [
{
"id": "gemini-2.5-flash",
"name": "Gemini 2.5 Flash",
"description": "Fast multimodal model via Vertex",
"descriptionKey": "models.descriptions.enterprise.vertex_gemini_2_5_flash"
},
{
"id": "gemini-2.5-pro",
"name": "Gemini 2.5 Pro",
"description": "Advanced reasoning via Vertex",
"descriptionKey": "models.descriptions.enterprise.vertex_gemini_2_5_pro"
}
]
}
],
"localProviders": [
{
"id": "qwen",
"name": "Qwen",
"baseUrl": "https://huggingface.co",
"models": [
{
"id": "qwen3.5-9b-q4_k_m",
"name": "Qwen3.5 9B",
"size": "5.5GB",
"sizeBytes": 5889811552,
"description": "Powerful hybrid model",
"fileName": "Qwen_Qwen3.5-9B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 262144,
"hfRepo": "bartowski/Qwen_Qwen3.5-9B-GGUF",
"recommended": true,
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_5_9b_q4_k_m"
},
{
"id": "qwen3.5-4b-q4_k_m",
"name": "Qwen3.5 4B",
"size": "2.7GB",
"sizeBytes": 2871743520,
"description": "Balanced Qwen3.5 for general use",
"fileName": "Qwen_Qwen3.5-4B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 262144,
"hfRepo": "bartowski/Qwen_Qwen3.5-4B-GGUF",
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_5_4b_q4_k_m"
},
{
"id": "qwen3.5-2b-q4_k_m",
"name": "Qwen3.5 2B",
"size": "1.3GB",
"sizeBytes": 1329766560,
"description": "Small Qwen3.5 for quick tasks",
"fileName": "Qwen_Qwen3.5-2B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 262144,
"hfRepo": "bartowski/Qwen_Qwen3.5-2B-GGUF",
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_5_2b_q4_k_m"
},
{
"id": "qwen3-8b-q4_k_m",
"name": "Qwen3 8B",
"size": "5.0GB",
"sizeBytes": 5402263552,
"description": "Latest Qwen3 with thinking mode support",
"fileName": "Qwen3-8B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "Qwen/Qwen3-8B-GGUF",
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_8b_q4_k_m"
},
{
"id": "qwen3-8b-q5_k_m",
"name": "Qwen3 8B (Q5)",
"size": "5.9GB",
"sizeBytes": 6281625600,
"description": "Higher quality Qwen3 with thinking mode",
"fileName": "Qwen3-8B-Q5_K_M.gguf",
"quantization": "q5_k_m",
"contextLength": 131072,
"hfRepo": "Qwen/Qwen3-8B-GGUF",
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_8b_q5_k_m"
},
{
"id": "qwen3-4b-q4_k_m",
"name": "Qwen3 4B",
"size": "2.5GB",
"sizeBytes": 2684354560,
"description": "Compact Qwen3 with reasoning capabilities",
"fileName": "Qwen3-4B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "Qwen/Qwen3-4B-GGUF",
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_4b_q4_k_m"
},
{
"id": "qwen3-1.7b-q8_0",
"name": "Qwen3 1.7B",
"size": "1.8GB",
"sizeBytes": 1965555712,
"description": "Small but capable Qwen3 model",
"fileName": "Qwen3-1.7B-Q8_0.gguf",
"quantization": "q8_0",
"contextLength": 131072,
"hfRepo": "Qwen/Qwen3-1.7B-GGUF",
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_1_7b_q8_0"
},
{
"id": "qwen3-32b-q4_k_m",
"name": "Qwen3 32B",
"size": "19.8GB",
"sizeBytes": 21260251955,
"description": "Most powerful local Qwen3 with deep reasoning",
"fileName": "Qwen3-32B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "Qwen/Qwen3-32B-GGUF",
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.qwen_qwen3_32b_q4_k_m"
},
{
"id": "qwen2.5-1.5b-instruct-q5_k_m",
"name": "Qwen2.5 1.5B",
"size": "1.3GB",
"sizeBytes": 1395864371,
"description": "Small model, good for basic tasks",
"fileName": "qwen2.5-1.5b-instruct-q5_k_m.gguf",
"quantization": "q5_k_m",
"contextLength": 32768,
"hfRepo": "Qwen/Qwen2.5-1.5B-Instruct-GGUF",
"descriptionKey": "models.descriptions.local.qwen_qwen2_5_1_5b_instruct_q5_k_m"
},
{
"id": "qwen2.5-3b-instruct-q5_k_m",
"name": "Qwen2.5 3B",
"size": "2.4GB",
"sizeBytes": 2620055552,
"description": "Balanced model for general use",
"fileName": "qwen2.5-3b-instruct-q5_k_m.gguf",
"quantization": "q5_k_m",
"contextLength": 32768,
"hfRepo": "Qwen/Qwen2.5-3B-Instruct-GGUF",
"descriptionKey": "models.descriptions.local.qwen_qwen2_5_3b_instruct_q5_k_m"
},
{
"id": "qwen2.5-7b-instruct-q4_k_m",
"name": "Qwen2.5 7B",
"size": "4.7GB",
"sizeBytes": 5025128858,
"description": "Large model with high quality (Q4_K_M)",
"fileName": "Qwen2.5-7B-Instruct-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 128000,
"hfRepo": "bartowski/Qwen2.5-7B-Instruct-GGUF",
"descriptionKey": "models.descriptions.local.qwen_qwen2_5_7b_instruct_q4_k_m"
},
{
"id": "qwen2.5-7b-instruct-q5_k_m",
"name": "Qwen2.5 7B (Q5)",
"size": "5.4GB",
"sizeBytes": 5841530470,
"description": "Large model, high quality reasoning (Q5_K_M)",
"fileName": "Qwen2.5-7B-Instruct-Q5_K_M.gguf",
"quantization": "q5_k_m",
"contextLength": 128000,
"hfRepo": "bartowski/Qwen2.5-7B-Instruct-GGUF",
"descriptionKey": "models.descriptions.local.qwen_qwen2_5_7b_instruct_q5_k_m"
}
]
},
{
"id": "mistral",
"name": "Mistral",
"baseUrl": "https://huggingface.co",
"models": [
{
"id": "mistral-nemo-12b-instruct-q4_k_m",
"name": "Mistral Nemo 12B",
"size": "7.5GB",
"sizeBytes": 8028321792,
"description": "Mistral's 12B model, strong multilingual performance",
"fileName": "Mistral-Nemo-12B-Instruct.Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "mradermacher/Mistral-Nemo-12B-Instruct-GGUF",
"recommended": true,
"descriptionKey": "models.descriptions.local.mistral_nemo_12b_instruct_q4_k_m"
},
{
"id": "mistral-7b-instruct-v0.3-q4_k_m",
"name": "Mistral 7B Instruct v0.3",
"size": "4.4GB",
"sizeBytes": 4692635648,
"description": "Fast and efficient instruction model",
"fileName": "Mistral-7B-Instruct-v0.3-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 32768,
"hfRepo": "bartowski/Mistral-7B-Instruct-v0.3-GGUF",
"descriptionKey": "models.descriptions.local.mistral_mistral_7b_instruct_v0_3_q4_k_m"
},
{
"id": "mistral-7b-instruct-v0.3-q5_k_m",
"name": "Mistral 7B Instruct v0.3 (Q5)",
"size": "5.1GB",
"sizeBytes": 5519900672,
"description": "Higher quality instruction model",
"fileName": "Mistral-7B-Instruct-v0.3-Q5_K_M.gguf",
"quantization": "q5_k_m",
"contextLength": 32768,
"hfRepo": "bartowski/Mistral-7B-Instruct-v0.3-GGUF",
"descriptionKey": "models.descriptions.local.mistral_mistral_7b_instruct_v0_3_q5_k_m"
}
]
},
{
"id": "llama",
"name": "Meta Llama",
"baseUrl": "https://huggingface.co",
"models": [
{
"id": "llama-3.2-1b-instruct-q4_k_m",
"name": "Llama 3.2 1B",
"size": "0.8GB",
"sizeBytes": 847249408,
"description": "Tiny model for edge devices",
"fileName": "Llama-3.2-1B-Instruct-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "bartowski/Llama-3.2-1B-Instruct-GGUF",
"descriptionKey": "models.descriptions.local.llama_llama_3_2_1b_instruct_q4_k_m"
},
{
"id": "llama-3.2-3b-instruct-q4_k_m",
"name": "Llama 3.2 3B",
"size": "2.0GB",
"sizeBytes": 2168958976,
"description": "Small but capable multilingual model",
"fileName": "Llama-3.2-3B-Instruct-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "bartowski/Llama-3.2-3B-Instruct-GGUF",
"recommended": true,
"descriptionKey": "models.descriptions.local.llama_llama_3_2_3b_instruct_q4_k_m"
},
{
"id": "llama-3.1-8b-instruct-q4_k_m",
"name": "Llama 3.1 8B",
"size": "4.9GB",
"sizeBytes": 5282717696,
"description": "Powerful model with great performance",
"fileName": "Meta-Llama-3.1-8B-Instruct-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF",
"descriptionKey": "models.descriptions.local.llama_llama_3_1_8b_instruct_q4_k_m"
}
]
},
{
"id": "openai-oss",
"name": "OpenAI",
"baseUrl": "https://huggingface.co",
"models": [
{
"id": "gpt-oss-20b-mxfp4",
"name": "GPT-OSS 20B",
"size": "12.1GB",
"sizeBytes": 12999763968,
"description": "OpenAI's open-weight model for consumer hardware",
"fileName": "gpt-oss-20b-MXFP4.gguf",
"quantization": "mxfp4",
"contextLength": 128000,
"hfRepo": "ggml-org/gpt-oss-20b-GGUF",
"recommended": true,
"supportsThinking": true,
"descriptionKey": "models.descriptions.local.openai_oss_gpt_oss_20b_mxfp4"
}
]
},
{
"id": "gemma",
"name": "Gemma (Google)",
"baseUrl": "https://huggingface.co",
"models": [
{
"id": "gemma-4-31b-it-q4_k_m",
"name": "Gemma 4 31B",
"size": "19.6GB",
"sizeBytes": 19598487712,
"description": "Google's largest Gemma 4, best quality",
"fileName": "google_gemma-4-31B-it-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 262144,
"hfRepo": "bartowski/google_gemma-4-31B-it-GGUF",
"descriptionKey": "models.descriptions.local.gemma_gemma_4_31b_it_q4_k_m"
},
{
"id": "gemma-4-31b-it-qat-q4_0",
"name": "Gemma 4 31B QAT",
"size": "17.9GB",
"sizeBytes": 17651001568,
"description": "Google's official QAT build, smaller with quality preserved",
"fileName": "gemma-4-31B_q4_0-it.gguf",
"quantization": "q4_0",
"contextLength": 262144,
"hfRepo": "google/gemma-4-31B-it-qat-q4_0-gguf",
"draftHfRepo": "unsloth/gemma-4-31B-it-qat-GGUF",
"draftFileName": "mtp-gemma-4-31B-it.gguf",
"draftSizeBytes": 279955968,
"descriptionKey": "models.descriptions.local.gemma_gemma_4_31b_it_qat_q4_0"
},
{
"id": "gemma-4-26b-a4b-it-q4_k_m",
"name": "Gemma 4 26B MoE",
"size": "17.0GB",
"sizeBytes": 17035037632,
"description": "Google's MoE Gemma 4, fast with 4B active params",
"fileName": "google_gemma-4-26B-A4B-it-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 262144,
"hfRepo": "bartowski/google_gemma-4-26B-A4B-it-GGUF",
"descriptionKey": "models.descriptions.local.gemma_gemma_4_26b_a4b_it_q4_k_m"
},
{
"id": "gemma-4-26b-a4b-it-qat-q4_0",
"name": "Gemma 4 26B MoE QAT",
"size": "14.7GB",
"sizeBytes": 14439363584,
"description": "Google's official QAT MoE build, smaller with quality preserved",
"fileName": "gemma-4-26B_q4_0-it.gguf",
"quantization": "q4_0",
"contextLength": 262144,
"hfRepo": "google/gemma-4-26B-A4B-it-qat-q4_0-gguf",
"draftHfRepo": "unsloth/gemma-4-26B-A4B-it-qat-GGUF",
"draftFileName": "mtp-gemma-4-26B-A4B-it.gguf",
"draftSizeBytes": 251939328,
"descriptionKey": "models.descriptions.local.gemma_gemma_4_26b_a4b_it_qat_q4_0"
},
{
"id": "gemma-4-e4b-it-q4_k_m",
"name": "Gemma 4 E4B",
"size": "5.4GB",
"sizeBytes": 5812137984,
"description": "Google's latest multimodal, strong reasoning",
"fileName": "google_gemma-4-E4B-it-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 262144,
"hfRepo": "bartowski/google_gemma-4-E4B-it-GGUF",
"descriptionKey": "models.descriptions.local.gemma_gemma_4_e4b_it_q4_k_m"
},
{
"id": "gemma-4-e4b-it-qat-q4_0",
"name": "Gemma 4 E4B QAT",
"size": "5.2GB",
"sizeBytes": 5154941280,
"description": "Google's official QAT build, strong reasoning, smaller size",
"fileName": "gemma-4-E4B_q4_0-it.gguf",
"quantization": "q4_0",
"contextLength": 262144,
"hfRepo": "google/gemma-4-E4B-it-qat-q4_0-gguf",
"draftHfRepo": "unsloth/gemma-4-E4B-it-qat-GGUF",
"draftFileName": "mtp-gemma-4-E4B-it.gguf",
"draftSizeBytes": 59678016,
"recommended": true,
"descriptionKey": "models.descriptions.local.gemma_gemma_4_e4b_it_qat_q4_0"
},
{
"id": "gemma-4-e2b-it-q4_k_m",
"name": "Gemma 4 E2B",
"size": "3.5GB",
"sizeBytes": 3711514624,
"description": "Google's compact Gemma 4, fast and efficient",
"fileName": "google_gemma-4-E2B-it-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 262144,
"hfRepo": "bartowski/google_gemma-4-E2B-it-GGUF",
"descriptionKey": "models.descriptions.local.gemma_gemma_4_e2b_it_q4_k_m"
},
{
"id": "gemma-4-e2b-it-qat-q4_0",
"name": "Gemma 4 E2B QAT",
"size": "3.4GB",
"sizeBytes": 3349516256,
"description": "Google's official compact QAT build, quality preserved",
"fileName": "gemma-4-E2B_q4_0-it.gguf",
"quantization": "q4_0",
"contextLength": 262144,
"hfRepo": "google/gemma-4-E2B-it-qat-q4_0-gguf",
"draftHfRepo": "unsloth/gemma-4-E2B-it-qat-GGUF",
"draftFileName": "mtp-gemma-4-E2B-it.gguf",
"draftSizeBytes": 59235648,
"descriptionKey": "models.descriptions.local.gemma_gemma_4_e2b_it_qat_q4_0"
},
{
"id": "gemma-3-12b-it-q4_k_m",
"name": "Gemma 3 12B",
"size": "7.3GB",
"sizeBytes": 7847624704,
"description": "Google's mid-range model, strong reasoning",
"fileName": "google_gemma-3-12b-it-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "bartowski/google_gemma-3-12b-it-GGUF",
"descriptionKey": "models.descriptions.local.gemma_gemma_3_12b_it_q4_k_m"
},
{
"id": "gemma-3-4b-it-q4_k_m",
"name": "Gemma 3 4B",
"size": "2.49GB",
"sizeBytes": 2489758112,
"description": "Google's open-weight model, great balance of speed and quality",
"fileName": "google_gemma-3-4b-it-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 131072,
"hfRepo": "bartowski/google_gemma-3-4b-it-GGUF",
"descriptionKey": "models.descriptions.local.gemma_gemma_3_4b_it_q4_k_m"
},
{
"id": "gemma-3-1b-it-q4_k_m",
"name": "Gemma 3 1B",
"size": "0.81GB",
"sizeBytes": 806058496,
"description": "Ultra-fast Google model, best for short dictation cleanup",
"fileName": "google_gemma-3-1b-it-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 32768,
"hfRepo": "bartowski/google_gemma-3-1b-it-GGUF",
"recommended": false,
"descriptionKey": "models.descriptions.local.gemma_gemma_3_1b_it_q4_k_m"
}
]
},
{
"id": "liquidai",
"name": "Liquid AI",
"baseUrl": "https://huggingface.co",
"models": [
{
"id": "lfm2.5-1.2b-instruct-q4_k_m",
"name": "LFM2.5 1.2B",
"size": "0.73GB",
"sizeBytes": 730895168,
"description": "Liquid AI's efficient on-device model, excellent instruction following",
"fileName": "LFM2.5-1.2B-Instruct-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 32768,
"hfRepo": "LiquidAI/LFM2.5-1.2B-Instruct-GGUF",
"recommended": true,
"descriptionKey": "models.descriptions.local.liquidai_lfm2_5_1_2b_instruct_q4_k_m"
},
{
"id": "lfm2.5-8b-a1b-q4_k_m",
"name": "LFM2.5 8B MoE",
"size": "4.8GB",
"sizeBytes": 5155564768,
"description": "Liquid AI's MoE, high quality with only 1.5B active params",
"fileName": "LFM2.5-8B-A1B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 32768,
"hfRepo": "LiquidAI/LFM2.5-8B-A1B-GGUF",
"descriptionKey": "models.descriptions.local.liquidai_lfm2_5_8b_a1b_q4_k_m"
},
{
"id": "lfm2-2.6b-q4_k_m",
"name": "LFM2 2.6B",
"size": "1.5GB",
"sizeBytes": 1563668704,
"description": "Balanced LFM2 for higher quality cleanup",
"fileName": "LFM2-2.6B-Q4_K_M.gguf",
"quantization": "q4_k_m",
"contextLength": 32768,
"hfRepo": "LiquidAI/LFM2-2.6B-GGUF",
"descriptionKey": "models.descriptions.local.liquidai_lfm2_2_6b_q4_k_m"
},
{
"id": "lfm2.5-350m-q8_0",
"name": "LFM2.5 350M",
"size": "0.38GB",
"sizeBytes": 379217632,
"description": "Ultra-light model for fast dictation cleanup",
"fileName": "LFM2.5-350M-Q8_0.gguf",
"quantization": "q8_0",
"contextLength": 32768,
"hfRepo": "LiquidAI/LFM2.5-350M-GGUF",
"descriptionKey": "models.descriptions.local.liquidai_lfm2_5_350m_q8_0"
},
{
"id": "lfm2.5-230m-q8_0",
"name": "LFM2.5 230M",
"size": "0.25GB",
"sizeBytes": 246598496,
"description": "Tiny model, instant cleanup on modest hardware",
"fileName": "LFM2.5-230M-Q8_0.gguf",
"quantization": "q8_0",
"contextLength": 32768,
"hfRepo": "LiquidAI/LFM2.5-230M-GGUF",
"descriptionKey": "models.descriptions.local.liquidai_lfm2_5_230m_q8_0"
}
]
}
],
"openwhisprCloudModels": {
"tiers": [
{
"id": "fast",
"label": "Fast",
"provider": "groq",
"models": [
{
"id": "llama-3.3-70b-versatile",
"name": "Llama 3.3 70B",
"description": "Fast, high-quality reasoning",
"recommended": true
},
{
"id": "llama-3.1-8b-instant",
"name": "Llama 3.1 8B",
"description": "Ultra-fast, low latency"
}
]
},
{
"id": "balanced",
"label": "Balanced",
"provider": "openrouter",
"models": [
{
"id": "anthropic/claude-sonnet-4.6",
"name": "Claude Sonnet 4.6",
"description": "Balanced performance and quality",
"recommended": true
},
{
"id": "google/gemini-2.5-flash",
"name": "Gemini 2.5 Flash",
"description": "Fast with strong reasoning"
}
]
},
{
"id": "quality",
"label": "Quality",
"provider": "openrouter",
"models": [
{
"id": "anthropic/claude-opus-4.8",
"name": "Claude Opus 4.8",
"description": "Most capable, highest quality",
"recommended": true
},
{
"id": "openai/gpt-4.1",
"name": "GPT-4.1",
"description": "Strong general-purpose reasoning"
}
]
}
]
}
}