Turn audio into subtitles with AI
Transcribe an authorised recording into timed segments and separate subtitle files. OpenClip can label speaker changes and deliver JSON, SRT and WebVTT outputs.
To turn audio into subtitles with AI, choose a recording you can upload and a format your player accepts. A short lesson may need a sidecar file for the video, while an editor checking quotations may want a transcript with a start and end time for every segment.
The openclip-transcription skill submits the file to OpenClip and waits for separate JSON, SRT and WebVTT results. Speaker labelling can distinguish turns in a conversation. It does not establish who a speaker is. The result links are unsigned and do not expire, which matters for interviews and private speech.
What it does
- Produce an SRT track for a recorded lesson.
- Choose WebVTT for a player that accepts a separate web caption file.
- Label different speaking turns in an interview transcript.
- Find the timestamp of a quotation and check it against the recording.
- Specify a known language when a short or noisy sample could confuse automatic detection.
Install and use
- Install the skill and authorise the official OpenClip MCP connection. The standalone feature uses an account with daily and per-file limits.
- Ask, “Transcribe this lesson, use its known language and return a separate subtitle track.”
- Review uncertain words and cue timings in your player. Use material suitable for output through permanent, unsigned links.
FAQ
Source
Made by OpenClip (Oak Ridge Ventures). Original skill: openclip-transcription, published on GitHub under the MIT licence.
View on GitHub