FunASR-powered video transcription, subtitle generation, and LLM-assisted clipping tool with a local Gradio UI.
Python ai-toolsai-video-editingasrauto-subtitleschinesecontent-creationfunasrfunclipgradiollmparaformerspeech-recognitionspeech-to-textsubtitles-generatortranscriptionvideo-editingvideo-processingvideo-subtitlesvideo-transcriptionwhisper-alternative
feat(llm): add Cheaper Inference as a named OpenAI-compatible gateway (#223)
aiapienthusiast36acf72e2add
diff --git a/README.md b/README.mdindex 65ded3f..c6ba6b4 100644--- a/README.md+++ b/README.md@@ -138,6 +138,12 @@ OrcaRouter exposes one endpoint for all frontier and open-weight models, so youSet `ORCAROUTER_API_KEY` (and optionally `ORCAROUTER_API_BASE`, which defaults to `https://api.orcarouter.ai/v1`) instead of pasting the key into the UI if you prefer environment-based configuration. A key is available at https://www.orcarouter.ai.+#### Using Cheaper Inference as your LLM gateway (optional)++FunClip can also send LLM-assisted clipping through [Cheaper Inference](https://cheaperinference.com), an OpenAI-compatible LLM gateway. One API key gives access to models from several labs. Select any `cheaperinference/` model in the **LLM Model Name** dropdown, paste a Cheaper Inference API key in the **APIKEY** box, and click 'LLM Inference'. FunClip removes the `cheaperinference/` prefix and sends the transcript and prompts to `https://api.cheaperinference.com/v1/chat/completions`. The returned segments work with the existing 'AI Clip' button.++Set `CHEAPER_INFERENCE_API_KEY` (and optionally `CHEAPER_INFERENCE_API_BASE`, which defaults to `https://api.cheaperinference.com/v1`) instead of pasting the key into the UI if you prefer environment-based configuration. A key is available at https://cheaperinference.com/signup.+#### Content-aware clipping with TwelveLabs Pegasus (optional)Besides the transcript-based LLMs above, FunClip can optionally use [TwelveLabs](https://twelvelabs.io) Pegasus, a video understanding model that reasons over the actual video (visuals + audio) rather than only the ASR transcript. This helps pick highlight segments even when the transcript alone is ambiguous (e.g. action, scene changes, on-screen events). To use it, select the `pegasus1.5` model name, paste your TwelveLabs API key, upload a video, and click 'LLM Inference' — Pegasus returns segments in the same `N. [start-end] text` format, so the existing 'AI Clip' button works unchanged. It needs `pip install twelvelabs`, and a free API key is available at https://twelvelabs.io.diff --git a/README_zh.md b/README_zh.mdindex 49a534a..7836eac 100644--- a/README_zh.md+++ b/README_zh.md@@ -163,6 +163,12 @@ OrcaRouter 用单一端点接入所有前沿与开源模型,无需修改 FunCl也可以不填 UI,而是设置 `ORCAROUTER_API_KEY` 环境变量(可选 `ORCAROUTER_API_BASE`,默认为 `https://api.orcarouter.ai/v1`)。Key 可在 https://www.orcarouter.ai 获取。+#### 使用 Cheaper Inference 作为 LLM 网关(可选)++FunClip 也可以将 LLM 智能裁剪发送到 [Cheaper Inference](https://cheaperinference.com)——一个 OpenAI 兼容的 LLM 网关。一个 API key 即可使用多家厂商的模型。在 **LLM Model Name** 下拉框选择任意 `cheaperinference/` 模型,在 **APIKEY** 输入框粘贴 Cheaper Inference API key,点击“LLM推理”。FunClip 会去掉 `cheaperinference/` 前缀,把字幕与 prompt 发送到 `https://api.cheaperinference.com/v1/chat/completions`,返回的分段与现有“AI Clip”按钮兼容。++也可以不填 UI,而是设置 `CHEAPER_INFERENCE_API_KEY` 环境变量(可选 `CHEAPER_INFERENCE_API_BASE`,默认为 `https://api.cheaperinference.com/v1`)。Key 可在 https://cheaperinference.com/signup 获取。+### B.通过命令行调用使用FunClip的相关功能```shell# 下载下面命令用到的示例视频diff --git a/funclip/launch.py b/funclip/launch.pyindex cfca30d..08ae700 100644--- a/funclip/launch.py+++ b/funclip/launch.py@@ -150,7 +150,7 @@ if __name__ == "__main__":)def llm_inference(system_content, user_content, srt_text, model, apikey, video_input=None):- SUPPORT_LLM_PREFIX = ['litellm', 'qwen', 'gpt', 'g4f', 'moonshot', 'deepseek', 'atlascloud', 'minimax', 'orcarouter', 'pegasus']+ SUPPORT_LLM_PREFIX = ['litellm', 'qwen', 'gpt', 'g4f', 'moonshot', 'deepseek', 'atlascloud', 'minimax', 'orcarouter', 'cheaperinference', 'pegasus']if model.startswith('litellm/'):return litellm_call(apikey, model, user_content+'\n'+srt_text, system_content)if model.startswith('pegasus'):@@ -162,7 +162,7 @@ if __name__ == "__main__":return call_twelvelabs_pegasus(apikey, video_input, model=model, prompt=system_content)if model.startswith('qwen'):return call_qwen_model(apikey, model, user_content+'\n'+srt_text, system_content)- if model.startswith('gpt') or model.startswith('moonshot') or model.startswith('deepseek') or model.startswith('atlascloud/') or model.startswith('minimax/') or model.startswith('orcarouter/'):+ if model.startswith('gpt') or model.startswith('moonshot') or model.startswith('deepseek') or model.startswith('atlascloud/') or model.startswith('minimax/') or model.startswith('orcarouter/') or model.startswith('cheaperinference/'):return openai_call(apikey, model, user_content+'\n'+srt_text, system_content)elif model.startswith('g4f'):model = "-".join(model.split('-')[1:])@@ -273,6 +273,10 @@ if __name__ == "__main__":"orcarouter/fusion","orcarouter/fusion-flash","orcarouter/fusion-mini",+ "cheaperinference/gpt-5.4-mini",+ "cheaperinference/gpt-5.4",+ "cheaperinference/claude-sonnet-5",+ "cheaperinference/deepseek-v4-flash","pegasus1.5"],value="deepseek-chat",label="LLM Model Name",diff --git a/funclip/llm/openai_api.py b/funclip/llm/openai_api.pyindex 4882558..66615df 100644--- a/funclip/llm/openai_api.py+++ b/funclip/llm/openai_api.py@@ -19,6 +19,11 @@ MINIMAX_MODEL_PREFIX = "minimax/"ORCAROUTER_API_BASE = "https://api.orcarouter.ai/v1"ORCAROUTER_MODEL_PREFIX = "orcarouter/"+# Cheaper Inference is an OpenAI-compatible LLM gateway. Its model IDs are+# bare (e.g. `gpt-5.4-mini`), so the `cheaperinference/` prefix is stripped.+CHEAPER_INFERENCE_API_BASE = "https://api.cheaperinference.com/v1"+CHEAPER_INFERENCE_MODEL_PREFIX = "cheaperinference/"+def _resolve_model_config(model):base_url = None@@ -53,6 +58,16 @@ def _resolve_model_config(model):if not base_url:base_url = ORCAROUTER_API_BASEapi_key_env = "ORCAROUTER_API_KEY"+ elif model.startswith(CHEAPER_INFERENCE_MODEL_PREFIX):+ model = model[len(CHEAPER_INFERENCE_MODEL_PREFIX):]+ if not model:+ raise ValueError(+ "Model name is empty after stripping cheaperinference/ prefix"+ )+ base_url = os.environ.get("CHEAPER_INFERENCE_API_BASE", CHEAPER_INFERENCE_API_BASE).strip()+ if not base_url:+ base_url = CHEAPER_INFERENCE_API_BASE+ api_key_env = "CHEAPER_INFERENCE_API_KEY"elif model.startswith("deepseek"):base_url = "https://api.deepseek.com"elif model.startswith("gpt-3.5-turbo"):diff --git a/tests/test_cheaperinference_api.py b/tests/test_cheaperinference_api.pynew file mode 100644index 0000000..f142f9c--- /dev/null+++ b/tests/test_cheaperinference_api.py@@ -0,0 +1,92 @@+"""Tests for Cheaper Inference routing through the OpenAI-compatible client."""++import os+import unittest+from unittest.mock import MagicMock, patch++from funclip.llm.openai_api import (+ CHEAPER_INFERENCE_API_BASE,+ openai_call,+)+++def _mock_completion(content="ok"):+ completion = MagicMock()+ completion.choices = [MagicMock()]+ completion.choices[0].message.content = content+ return completion+++class TestCheaperInferenceRouting(unittest.TestCase):+ def test_cheaperinference_prefix_uses_gateway_base_url(self):+ client = MagicMock()+ client.chat.completions.create.return_value = _mock_completion("clip plan")++ with patch("funclip.llm.openai_api.OpenAI", return_value=client) as openai_cls:+ result = openai_call(+ "ci-key",+ "cheaperinference/gpt-5.4-mini",+ "subtitle text",+ "find highlights",+ )++ self.assertEqual(result, "clip plan")+ openai_cls.assert_called_once_with(+ api_key="ci-key",+ base_url=CHEAPER_INFERENCE_API_BASE,+ )+ # Cheaper Inference model IDs are bare: the prefix is stripped.+ call_kwargs = client.chat.completions.create.call_args[1]+ self.assertEqual(call_kwargs["model"], "gpt-5.4-mini")++ def test_cheaperinference_api_key_falls_back_to_env(self):+ client = MagicMock()+ client.chat.completions.create.return_value = _mock_completion()++ with patch.dict(os.environ, {"CHEAPER_INFERENCE_API_KEY": "env-ci-key"}, clear=False):+ with patch("funclip.llm.openai_api.OpenAI", return_value=client) as openai_cls:+ openai_call("", "cheaperinference/gpt-5.4-mini", "text")++ openai_cls.assert_called_once_with(+ api_key="env-ci-key",+ base_url=CHEAPER_INFERENCE_API_BASE,+ )+ call_kwargs = client.chat.completions.create.call_args[1]+ self.assertEqual(call_kwargs["model"], "gpt-5.4-mini")++ def test_cheaperinference_api_base_env_overrides(self):+ client = MagicMock()+ client.chat.completions.create.return_value = _mock_completion()++ with patch.dict(+ os.environ,+ {"CHEAPER_INFERENCE_API_BASE": "https://gateway.example.com/v1"},+ clear=False,+ ):+ with patch("funclip.llm.openai_api.OpenAI", return_value=client) as openai_cls:+ openai_call("ci-key", "cheaperinference/claude-sonnet-5", "text")++ openai_cls.assert_called_once_with(+ api_key="ci-key",+ base_url="https://gateway.example.com/v1",+ )++ def test_empty_cheaperinference_model_raises(self):+ with self.assertRaises(ValueError):+ openai_call("key", "cheaperinference/", "text")++ def test_missing_cheaperinference_key_does_not_fall_back_to_openai_key(self):+ with patch.dict(+ os.environ,+ {"OPENAI_API_KEY": "openai-only-key"},+ clear=True,+ ):+ with patch("funclip.llm.openai_api.OpenAI") as openai_cls:+ with self.assertRaisesRegex(ValueError, "CHEAPER_INFERENCE_API_KEY"):+ openai_call("", "cheaperinference/gpt-5.4-mini", "text")++ openai_cls.assert_not_called()+++if __name__ == "__main__":+ unittest.main()diff --git a/tests/test_cheaperinference_launch_integration.py b/tests/test_cheaperinference_launch_integration.pynew file mode 100644index 0000000..03d1f52--- /dev/null+++ b/tests/test_cheaperinference_launch_integration.py@@ -0,0 +1,74 @@+"""Regression tests for Cheaper Inference choices and prompt routing in the launcher."""++import ast+import unittest+from pathlib import Path+++LAUNCH_PATH = Path(__file__).resolve().parents[1] / "funclip" / "launch.py"+++class TestCheaperInferenceLaunchIntegration(unittest.TestCase):+ @classmethod+ def setUpClass(cls):+ cls.tree = ast.parse(LAUNCH_PATH.read_text(encoding="utf-8"))++ def test_openai_compatible_route_handles_cheaperinference_prefix(self):+ llm_inference = next(+ node+ for node in ast.walk(self.tree)+ if isinstance(node, ast.FunctionDef) and node.name == "llm_inference"+ )+ openai_call = next(+ node+ for node in ast.walk(llm_inference)+ if isinstance(node, ast.Call)+ and isinstance(node.func, ast.Name)+ and node.func.id == "openai_call"+ )+ # The openai_call dispatch branch must also cover cheaperinference/ models.+ dispatch_condition = next(+ node+ for node in ast.walk(llm_inference)+ if isinstance(node, ast.If) and isinstance(node.test, ast.BoolOp)+ )+ self.assertIn("cheaperinference/", ast.unparse(dispatch_condition.test))+ self.assertEqual(ast.unparse(openai_call.args[2]), "user_content + '\\n' + srt_text")+ self.assertEqual(ast.unparse(openai_call.args[3]), "system_content")++ def test_support_prefix_list_includes_cheaperinference(self):+ llm_inference = next(+ node+ for node in ast.walk(self.tree)+ if isinstance(node, ast.FunctionDef) and node.name == "llm_inference"+ )+ assigned = [+ node+ for node in ast.walk(llm_inference)+ if isinstance(node, ast.Assign) and node.targets[0].id == "SUPPORT_LLM_PREFIX"+ ]+ self.assertEqual(len(assigned), 1)+ prefix_list = assigned[0].value+ self.assertIsInstance(prefix_list, ast.List)+ prefixes = {+ node.value+ for node in ast.walk(prefix_list)+ if isinstance(node, ast.Constant) and isinstance(node.value, str)+ }+ self.assertIn("cheaperinference", prefixes)++ def test_dropdown_lists_cheaperinference_gateway_models(self):+ string_literals = {+ node.value+ for node in ast.walk(self.tree)+ if isinstance(node, ast.Constant) and isinstance(node.value, str)+ }++ self.assertIn("cheaperinference/gpt-5.4-mini", string_literals)+ self.assertIn("cheaperinference/gpt-5.4", string_literals)+ self.assertIn("cheaperinference/claude-sonnet-5", string_literals)+ self.assertIn("cheaperinference/deepseek-v4-flash", string_literals)+++if __name__ == "__main__":+ unittest.main()