# modelscope/FunClip

FunASR-powered video transcription, subtitle generation, and LLM-assisted clipping tool with a local Gradio UI.

Repository: https://github.com/modelscope/FunClip
Canonical: https://ross.abutalabs.com/products/funclip
Homepage: https://huggingface.co/spaces/FunAudioLLM/FunClip
Language: Python
License: MIT
License Family: permissive
Topics: speech-recognition, video-subtitles, subtitles-generator, speech-to-text, gradio, llm, ai-tools, ai-video-editing, content-creation, video-editing, video-processing, asr, auto-subtitles, chinese, funasr, paraformer, transcription, whisper-alternative, funclip, video-transcription
Last push: 2026-08-19T02:27:44+00:00

## Health v2 (maintenance only)
Score: 95/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 98, release rhythm 96, longevity 86
- inputs: {"age_days": 1204, "days_push": 15, "days_rel": 30, "gap_med": 9, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6190, forks 742 (observed 2026-08-28T04:09:39.046121+00:00)

## What it is
FunClip is an open-source, locally deployed video clipping tool that uses FunASR Paraformer models for speech recognition and subtitle generation, with LLM-assisted segment selection. It provides a Gradio web UI for selecting text segments or speakers and exporting clipped videos with SRT subtitles.

## Use cases
- clip video segments by searching transcribed text
- auto-generate subtitles for videos
- cut clips from a specific speaker using speaker recognition
- use an LLM to pick the best moments to clip
- transcribe Chinese videos with hotword customization
- self-host a video clipping web app

## When to choose
- you need accurate Chinese ASR with timestamps for video editing
- you want fully local, open-source clipping without uploading footage
- you want to clip by speaker or by quoted transcript text
- you need SRT subtitle export for clips and full videos

## When to avoid
- you need polished non-Chinese-language ASR out of the box
- you want a turnkey cloud service rather than a local Python setup
- you need timeline-based manual video editing rather than text-driven clipping

## Facets
- artifact type: application
- maturity: active
- function: speech-recognition, video-processing, llm-inference, gui
- domain: speech-processing, media
- platform: python, cross-platform, self-hosted
- tags: asr, video-clipping, subtitles, gradio, paraformer, speaker-recognition, chinese, llm, content-creation, video, natural-language-processing, web-server

## Member repositories
- modelscope/FunClip (main) score 95

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:39.046121+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:47:22.493591+00:00, confidence not recorded.
  - readme: https://github.com/modelscope/FunClip (fetched 2026-08-28T04:09:39.046121+00:00, sha 39e3dd4e40fd)
  - homepage: https://huggingface.co/spaces/FunAudioLLM/FunClip (fetched 2026-08-29T08:43:52.405760+00:00, sha be969d2a1950)
- Data as of 2026-08-30T08:39:29.467469+00:00.