TOOLORA / Local reading edition
FunClip
Recognize video speech, select transcript passages or speaker segments, and export clips with SRT subtitles; LLM-assisted selection is optional.
Overview
FunClip uses recognized text, timing and speaker labels to edit video. Select passages or a speaker in Gradio, combine segments and export clips plus full or segment SRT subtitles. An optional configured LLM helps choose content. Default recognition, speaker models and the optional MOSS path are separate components; labels are anonymous, not real identities. Remote LLM access differs from local recognition. Code uses MIT, and model dependencies must be prepared separately.
Official sources and references
- Official websitehttps://github.com/modelscope/FunClip
Descriptions summarize public information; consult official websites for features, pricing and regional availability.