# Clips AI

An open-source Python library that helps developers turn longform video into clips and reframed formats.

- Findings24: <https://findings24.com/pt/tools/clips-ai>
- Site: <https://clipsai.com>
- Preços: Freemium
- Categorias: Ferramentas Dev IA, Reaproveitamento de vídeos com IA, Editor de Vídeo IA

## O que é Clips AI?

Clips AI is an open-source Python library for repurposing longform video. Developers can use it to identify segments from a video's transcript and create clips, then dynamically reframe video for alternate aspect ratios such as 9:16. Its workflow is aimed at audio-centric, narrative videos where spoken content and the active speaker guide editing.

## O que Clips AI pode fazer?

### Transcript-based clip discovery

Transcribes video audio with WhisperX and analyzes the transcript to identify candidate clip start and end times.

### Clip segmentation

Finds multiple segments in a long video, returning start and end times that can be used to trim the original footage.

### Dynamic reframing

Resizes video to a requested aspect ratio, such as 9:16, by generating crops that follow the current speaker.

### Speaker diarization for resizing

Uses Pyannote for speaker diarization as part of the video-resizing workflow; a Hugging Face access token is required.

### Python library workflow

Provides programmatic functionality through Python imports, including Transcriber, ClipFinder, and resize.

## Casos de uso

### Podcast highlight clips

Identify self-contained spoken segments from a long podcast recording for trimming into shorter clips.

### Interview excerpts

Use transcript-led segment discovery to locate answers or discussion sections worth publishing separately.

### Speech and sermon repurposing

Find passages in speeches or sermons and prepare them as individual clips from a longer recording.

### Vertical video versions

Reframe existing footage to a portrait aspect ratio, with crops dynamically focused on the active speaker.

## Perguntas frequentes

### What kinds of videos is Clips AI designed for?

It is designed for audio-centric, narrative videos, including podcasts, interviews, speeches, and sermons.

### How does clip discovery work?

The video must first be transcribed. The documented workflow uses WhisperX, then passes the transcription to ClipFinder to identify clip segments and their timestamps.

### Can it make vertical video?

Yes. The resize function accepts a target aspect ratio, with 9:16 shown as an example, and dynamically reframes around the current speaker.

### What is needed to resize video?

The resizing workflow requires a Pyannote access token obtained through Hugging Face for speaker diarization. The site states that Pyannote usage is not charged.

## Informações rápidas

- **Product type**: Open-source Python library
- **Primary users**: Developers
- **Input workflow**: Transcribe video audio before finding clips
- **Transcription component**: WhisperX
- **Resizing component**: Pyannote speaker diarization; Hugging Face access token required