analyze

From video to structured text, in one call.

Define the format in your prompt. Pegasus analyzes visuals, audio, and speech, then returns the exact structure you asked for: summaries, chapters, reports, time-coded metadata, or any JSON schema you specify.

What you can do with Pegasus.

Generate summaries, chapters, reports, and structured metadata from any video up to 2 hours long. Pegasus understands on-screen text alongside speech, so your output includes the whiteboard, the jersey number, and the slide.

Analysis across content type.

Prompt for any output structure: titles, Q&As, reports, JSON schemas, time-coded metadata. The format is yours to define.

Summarize any content.

Generate descriptions, highlights, and chapter markers from videos up to 2 hours long without pre-indexing. Send a URL, receive structured output.

Extract tags and insights.

Generate categories, hashtags, and time-coded metadata that flow directly into your search index, recommendation engine, or training pipeline.

What you can do with Pegasus.

Generate summaries, chapters, reports, and structured metadata from any video up to 2 hours long. Pegasus understands on-screen text alongside speech, so your output includes the whiteboard, the jersey number, and the slide.

Analysis across content type.

Prompt for any output structure: titles, Q&As, reports, JSON schemas, time-coded metadata. The format is yours to define.

Summarize any content.

Generate descriptions, highlights, and chapter markers from videos up to 2 hours long without pre-indexing. Send a URL, receive structured output.

Extract tags and insights.

Generate categories, hashtags, and time-coded metadata that flow directly into your search index, recommendation engine, or training pipeline.

Start building with Pegasus.

Any text format, from any video.

From Q&As and hashtag generation to detailed content reports.

Fast and accurate chapterization

Generate titles, headings, and chapters in seconds.

Text prompt generation

Index once, then generate boundless text prompts.

Content analysis and incident reporting

Get time-stamped verbal descriptions of events captured on footage.

Content summarization

Get immediate, high-quality summaries of any video.

Metadata generation

Move beyond manual tagging to instant metadata creation.

Start building with Pegasus.

Any text format, from any video.

From Q&As and hashtag generation to detailed content reports.

Fast and accurate chapterization

Generate titles, headings, and chapters in seconds.

Text prompt generation

Index once, then generate boundless text prompts.

Content analysis and incident reporting

Get time-stamped verbal descriptions of events captured on footage.

Content summarization

Get immediate, high-quality summaries of any video.

Metadata generation

Move beyond manual tagging to instant metadata creation.

Start building with Pegasus.

Any text format, from any video.

From Q&As and hashtag generation to detailed content reports.

Fast and accurate chapterization

Generate titles, headings, and chapters in seconds.

Text prompt generation

Index once, then generate boundless text prompts.

Content analysis and incident reporting

Get time-stamped verbal descriptions of events captured on footage.

Content summarization

Get immediate, high-quality summaries of any video.

Metadata generation

Move beyond manual tagging to instant metadata creation.

Python
Node.js
1import requests
2 
3# Step 2: Define the API URL and the specific endpoint
4API_URL = "https://api.twelvelabs.io/v1.3"
5INDEXES_URL = f"{API_URL}/indexes"
6 
7# Step 3: Create the necessary headers for authentication
8headers = {
9 "x-api-key": "<YOUR_API_KEY>"
10}
11 
12# Step 4: Prepare the data payload for your API request
13INDEX_NAME = "<YOUR_INDEX_NAME>"
14data = {
15 "models": [
16 {
17 "model_name": "marengo3.0",
18 "model_options": ["visual", "audio"]
19 }
20 ]
21}

Integrate with your personalized SDK — and your vision.

Do more with your video from day one with our easy APIs and developer-friendly SDKs. This is AI made to work for you, ready to integrate and adapt.

Integrate with your personalized SDK — and your vision.

Do more with your video from day one with our easy APIs and developer-friendly SDKs. This is AI made to work for you, ready to integrate and adapt.

Try it on Playground

Turn your video into something you can build with.