---
title: Multimodal Feature Extractor
emoji: 🎯
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: 5.9.1
app_file: app.py
pinned: false
license: mit
short_description: Extract embeddings from video, image, GIF, and text
---

# Mixpeek Multimodal Feature Extractor

Extract embeddings and features from **videos**, **images**, **GIFs**, and **text** using Mixpeek's multimodal AI pipeline.

## Features

- **Video Processing**: Decompose videos into segments using time, scene, or silence-based splitting
- **Image & GIF Support**: Direct embedding generation for static and animated images
- **Text Embedding**: Generate embeddings from plain text
- **Transcription**: Whisper-powered speech-to-text (95%+ accuracy)
- **OCR**: Extract text from video frames and images
- **AI Descriptions**: Generate natural language descriptions of content
- **1408D Embeddings**: Google Vertex AI multimodal vectors for semantic search

## Supported Formats

| Type | Formats |
|------|---------|
| Video | MP4, MOV, AVI, MKV, WebM, FLV |
| Image | JPG, PNG, WebP, BMP |
| GIF | Animated and static GIFs |
| Text | Plain text strings |

## Usage

1. Get your API key from [mixpeek.com](https://mixpeek.com)
2. Choose your input type (file upload or text)
3. Configure extraction settings
4. Click "Extract Features" to process

## Output

Each extracted segment contains:
- Timing data (start/end times for video segments)
- Transcription text
- 1408D multimodal embedding vector
- OCR extracted text (if enabled)
- AI-generated description (if enabled)
- Thumbnail URL (if enabled)

## Links

- [Mixpeek Documentation](https://docs.mixpeek.com)
- [API Reference](https://docs.mixpeek.com/api-reference)
- [Get API Key](https://mixpeek.com)