# api.video MCP for AI Agents MCP

> api.video allows you to manage every aspect of video content—from initial upload and live stream setup to advanced analytics and chapter creation—all through natural conversation with your AI client.

## Overview
- **Category:** developer-tools
- **Price:** Free
- **Tags:** video-hosting, live-streaming, video-api, encoding, video-analytics

## Description

This MCP gives your agent total control over complex video infrastructure. You can handle the entire lifecycle, starting by ingesting raw footage from a URL or uploading it directly. It lets you manage the container itself, set up custom player themes, and even schedule watermarks for protection. For live content, you can create streams, manage broadcast details, and eventually signal when the session is complete. Beyond storage, your agent pulls detailed metrics like viewership breakdowns and time-series data to analyze performance. You'll also get deep control over metadata by creating summaries or listing chapters. If you’re already using a platform like Vinkius, connecting this MCP gives your AI client all the video tools it needs in one place.

## Tools

### complete_live_stream
Stops a live stream that is currently running.

### create_live_stream
Creates a new object defining the parameters for a live broadcast.

### create_player
Generates and customizes a player theme, allowing you to set specific visual elements.

### create_summary
Automatically generates an abstract and key takeaways for a specified video content.

### create_upload_token
Generates a temporary token used to securely upload large files from client-side applications.

### create_video
Establishes a new video object container for tracking and management.

### create_webhook
Subscribes to specific video events, ensuring you are notified when something happens to your content.

### delete_caption
Removes a specific caption track from a video.

### delete_chapter
Deletes defined time segment markers (chapters) from a video timeline.

### delete_live_stream
Permanently removes an existing live stream definition.

### delete_player_logo
Removes the logo image from a custom player theme.

### delete_player
Deletes an entire customized player theme.

### delete_summary
Permanently removes generated summary text and takeaways for a video.

### delete_upload_token
Invalidates and deletes an active upload token.

### delete_video
Permanently removes a video object from your project container.

### delete_watermark
Removes a defined watermark overlay from the video content.

### delete_webhook
Unsubscribes and removes an existing webhook listener.

### get_caption
Retrieves the full text and timing data for a specific caption track.

### get_chapter
Fetches the details and timecodes for a specific chapter marker in a video.

### get_discarded_video
Retrieves historical data and details of a video that was discarded from the active project.

### get_live_stream
Fetches all current settings and status information for a specific live stream.

### get_metrics_breakdown
Pulls an analysis of viewership metrics, providing detailed breakdowns by source or time period.

### get_metrics_timeseries
Retrieves performance data points over a specific timeline for trend analysis.

### get_metrics
Gets an overall, aggregated view of all key analytics metrics for your content.

### get_player
Retrieves the full configuration and theme details of a specific player type.

### get_summary_source
Fetches the original text or source material used to generate a video summary.

### get_upload_token
Retrieves the details and expiry time of an active upload token.

### get_video_status
Checks if a video has finished uploading or encoding, providing its current readiness status.

### get_video
Retrieves all core details about a specific video object in the project.

### get_webhook
Fetches the configuration and security signature for an existing webhook listener.

### list_captions
Lists all available caption tracks associated with a video container.

### list_chapters
Retrieves a list of all defined chapters and their timecodes for a given video.

### list_discarded_videos
Lists videos that were discarded, if the Video Restore feature is enabled on your account.

### list_live_streams
Retrieves a list of all existing live stream definitions in the project.

### list_players
Lists every customized player theme that has been created for your videos.

### list_summaries
Retrieves a list of all generated summaries across your video library.

### list_tags
Shows every tag used in the project and how many videos are associated with each one.

### list_upload_tokens
Lists all active upload tokens to help manage your content pipeline security.

### list_videos
Retrieves a list of every video object currently tracked in the project container.

### list_watermarks
Lists all existing, defined watermarks used across your videos.

### list_webhooks
Shows a list of all configured webhook endpoints and their current status.

### restore_discarded_video
Moves a video from the discarded state back into active project management.

### update_caption_default
Sets one specific caption track to be displayed by default on the player.

### update_live_stream
Modifies existing live stream settings, such as titles or broadcast parameters.

### update_player
Makes changes to the visual appearance and settings of an existing player theme.

### update_summary_source
Allows manual editing or updating of the source text for a video summary.

### update_thumbnail
Changes the preview image shown for a video by selecting a specific timecode.

### update_video
Modifies general parameters of an entire video object, like its title or description.

### upload_video_source
Ingests new video content directly from a provided web URL.

## Prompt Examples

**Prompt:** 
```
List all videos I have, check which ones are ready to publish, and get their view count.
```

**Response:** 
```
**Video Inventory Report**

| Video Title | ID | Status | View Count |
| :--- | :--- | :--- | :--- |
| Product Demo Q1 | vi4k0j... | Ready (100%) | 8,902 |
| Interview with Sarah | vi2l9m... | Encoding (720p) | 1,550 |
| Tutorial: Basics | viabcde... | Complete | 22,101 |

Please review the details for 'Product Demo Q1' if you need to update its thumbnail.
```

**Prompt:** 
```
I need a new live stream for next week's conference. Can you set it up and give me the broadcast keys?
```

**Response:** 
```
**Live Stream Setup Complete**

✅ **Stream Title:** Annual Tech Summit 2024
✅ **Start Date:** Oct 15, 2024
✅ **RTMP Endpoint:** rtmp://broadcast.api.video/s
✅ **Stream Key:** `7f8e9a...` (Keep this private)

You can start broadcasting anytime using the key provided. I've also updated the player theme to match your new branding.
```

**Prompt:** 
```
Can you analyze the performance of my last five videos and give me a summary?
```

**Response:** 
```
Here is an aggregated view of the top 5 video performances:

*   **Tutorial: Basics:** Highest engagement. Viewership peaked on Tuesday afternoon.
*   **Product Demo Q1:** Strong retention rates, especially in the first 3 minutes.
*   **Interview with Sarah:** Steady viewership, suggesting good organic sharing.

I've generated a full text summary for each one and updated their metadata so you can reference key topics easily.
```

## Capabilities

### Manage Video Assets
Create new video containers, upload sources from URLs, and update core parameters for any existing video.

### Control Live Broadcasts
Set up live streams, manage stream keys, and terminate running broadcasts via simple commands.

### Analyze Performance Data
Retrieve aggregated metrics or detailed time-series breakdowns to understand how videos are performing for your audience.

### Enrich Video Metadata and Presentation
Generate automatic summaries, list chapters, add captions, and customize the appearance of players with watermarks.

### Automate Content Pipelines
Set up webhooks to react instantly when video events occur and generate upload tokens for secure client-side data ingestion.

## Use Cases

### Analyzing a Campaign's Reach
A marketing analyst needs to know which videos performed best last quarter. They ask their agent to use the MCP, which calls `get_metrics` and then runs `get_metrics_timeseries` to pull a detailed graph of viewership spikes over time.

### Updating Evergreen Content
A content manager noticed an old video is missing key chapters. They ask their agent, which uses the MCP's tools to `list_chapters` first, then executes a specific chapter update using `get_chapter`, fixing the metadata without touching the source file.

### Securing Upload Pipelines
A developer needs their client application to upload video files securely. They ask their agent to use `create_upload_token` and then monitor it with `get_upload_token`, ensuring the process is auditable and safe.

### Handling Live Broadcasts
A producer needs to set up a multi-day training event. They ask their agent to use `create_live_stream` for the initial setup, and later confirm the end by calling `complete_live_stream`, fully managing the live cycle.

## Benefits

- Get instant content analysis: Use the `create_summary` tool to generate abstracts and key takeaways for any video, eliminating manual transcription work.
- Maintain control over distribution: By running `create_player` and managing watermarks, you ensure your brand identity is consistently applied across all viewed videos.
- Streamline live broadcasting: The MCP allows you to use `create_live_stream` and later `complete_live_stream`, handling the entire lifecycle of a broadcast session.
- Deep analytics access: You don't just get numbers; you can call `get_metrics_breakdown` or `get_metrics_timeseries` to pull specific performance data points for reporting.
- Simplify content ingestion: Instead of complicated API calls, your agent handles uploading sources using `upload_video_source` and managing secure tokens with `create_upload_token`.

## How It Works

The bottom line is that you talk to your AI client about video tasks—like analyzing performance or updating chapters—and it handles all the necessary backend calls automatically.

1. Subscribe to this MCP in Vinkius and provide your api.video API Key.
2. Your AI agent accesses the full suite of video tools, allowing you to issue commands like 'List all videos' or 'Get metrics breakdown'.
3. The system executes the request against the live API connection, returning structured data that your agent translates back into actionable conversation.

## Frequently Asked Questions

**How do I use the api.video MCP to manage my live broadcasts?**
You tell your agent to create a new stream, and it handles all the necessary setup, providing you with the required broadcast keys. It also lets you update settings or signal when the stream is finished.

**Can I use this MCP to analyze my video performance data?**
Yes. You can ask your agent to pull detailed analytics. It retrieves both overall metrics and time-series breakdowns, helping you pinpoint exactly when and why viewers dropped off.

**Is this better than manually uploading videos via the website?**
Much better. Instead of multiple web logins, your agent handles everything—from ingesting a video source from a URL to checking its final encoding status—all within one chat session.

**How do I add chapters or captions using the api.video MCP?**
You simply ask for it. Your agent uses tools to list existing chapters, and then you can direct it to create new ones at specific timecodes, updating your video's metadata instantly.

**Can I customize the player appearance with this MCP?**
Absolutely. You tell your agent what changes you want—like adding a watermark or changing colors—and it uses tools to update the player theme and logo without needing developer access.