Practical overview
When it fits
Use ClipsAI when building a Python workflow around spoken, narrative video and you need transcript-derived clip candidates, aspect-ratio conversion, or both.
Who it is for
- Python developers building video-processing workflows
- Creators or teams working with podcasts, interviews, speeches or sermons
- Editors who want transcript-based candidate segments or speaker-aware reframing in an existing pipeline
Documented uses
- Identify candidate start and end times for short segments from a video's transcript.
- Reframe video into a target aspect ratio, including 16:9-to-9:16 conversion.
- Use speaker diarization to focus the crop on the current speaker.
Requirements & limits
The project is an open-source Python library under the MIT license. Its README lists installation through pip and requires a Hugging Face access token for the Pyannote-based resizing operation.
- Clip discovery depends on a transcript; the README's example uses WhisperX.
- ClipFinder returns candidate boundaries, while trimming the original video is described separately.
- Resizing uses Pyannote speaker diarization and requires a Hugging Face access token.
- The listed installation prerequisites include libmagic and ffmpeg, in addition to Python packages.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Find candidate clip times from a video transcript
Use WhisperX-backed transcription and ClipFinder to retrieve candidate start and end times.
Before you start
- Python
- ClipsAI and WhisperX installed
- libmagic and ffmpeg installed
- A video file accessible by path
Steps
Install ClipsAI with pip install clipsai.
Install WhisperX as shown in the README, and install libmagic and ffmpeg.
Use ClipsAI's Transcriber to transcribe the video, then pass the transcription to ClipFinder.find_clips.
Read the start_time and end_time values from the returned clip candidates.
Expected outcome
A set of candidate clip boundaries derived from the video's transcript.
Guide limitations
- This produces candidate times; the README points to a separate clipping reference for trimming the original video.
- The library is designed for audio-centric, narrative-based videos.
Resize a video to a target aspect ratio
Use ClipsAI's resize function to request a speaker-aware crop at a specified aspect ratio.
Before you start
- Python and ClipsAI installed
- libmagic and ffmpeg installed
- A video file accessible by path
- A Hugging Face access token for Pyannote speaker diarization
Steps
Install ClipsAI and the listed system prerequisites.
Obtain the Hugging Face access token required for the Pyannote-based diarization.
Call ClipsAI's resize function with the video path, token and desired aspect ratio, such as (9, 16).
Read the returned crop segments.
Expected outcome
Crop segments for reframing the video to the requested aspect ratio, with speaker-aware focus.
Guide limitations
- A Hugging Face access token is required for resizing.
- This guide covers the resize operation, not exporting a finished edited clip.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 16 October 2026
- Original website title
- ClipsAI/clipsai · Official homepage title · retained in its source language