Clips AI

Clips AI is an open-source Python library for converting longform videos into clips and changing video aspect ratios. Clips AI is designed for audio-centric, narrative-based recordings, including podcasts, interviews, speeches, and sermons. The documented approach combines transcript analysis for clip discovery with dynamic reframing that follows the current speaker.

Clips AI is used through Python code rather than a documented visual editing workflow. The website provides a quickstart, installation commands, and code examples for finding clips and calculating video crops. Separate clipping and resizing references cover trimming the original video and resizing footage to the desired aspect ratio.

Clips AI provides related capabilities for finding segments and adapting video framing. Transcript analysis identifies clip boundaries, while the resizing algorithm focuses the frame on the current speaker. The documentation demonstrates both workflows with Python examples and describes conversion from landscape footage into a portrait aspect ratio.

Clips AI installation involves installing the clipsai Python package, WhisperX, libmagic, and ffmpeg. The documentation recommends a virtual environment such as venv to avoid dependency conflicts. For clipping, the example creates a Transcriber, transcribes a video file, and passes the resulting transcription to ClipFinder to retrieve clip boundaries.

Clips AI resizing uses a video file path, a target aspect ratio, and a Pyannote authentication token. Pyannote provides speaker diarization, and a Hugging Face access token is required for that workflow. The example returns crop segments; the resizing reference explains resizing the original video to the desired aspect ratio.

Clips AI is intended for work with audio-centric videos whose content follows a narrative. The named examples are podcasts, interviews, speeches, and sermons. Because the documented workflows use Python installation commands and code, Clips AI fits users who can run Python and install the required video and transcription dependencies.

Clips AI is available as an open-source Python library, with installation instructions and a GitHub link on the documentation page. Access follows the documented package-installation workflow. The page does not list paid plans or subscription prices, and explicitly states that users will not be charged for using Pyannote.

Clips AI combines transcript-based clip discovery and speaker-focused reframing in a Python library. The documentation provides concrete setup commands and usage examples for both workflows. The main constraints are technical setup and prerequisites: clip finding depends on an existing transcription, while resizing requires Pyannote access through a Hugging Face token.