Pro Tier

Video Context MCP has a Pro tier that unlocks additional capabilities beyond the free defaults.

Pro features

Sensitive video redaction

The redact_sensitive tool blurs, pixelates, or blackouts rectangular regions in videos to hide API keys, tokens, passwords, and other secrets before sharing or analysis. It supports three operating modes:

Intent-driven — describe what to redact in natural language (e.g., "blur the API keys") and AI detection is enabled automatically
Manual mode — supply pixel coordinates directly; never sends frames to an AI provider
AI-assisted detection — samples frames and uses any of the 6 AI backends to detect sensitive regions

Requires a Pro license (VIDEO_MCP_LICENSE_KEY). See redact_sensitive documentation for full details.

Media generation (MiniMax)

All five media generation tools are exclusively available to Pro users:

Tool	Description
`text_to_speech`	Convert text to natural speech (MiniMax TTS)
`generate_image`	Generate images from a text prompt
`generate_video`	Generate video from text or image (async)
`generate_music`	Generate music from a prompt and optional lyrics
`query_generation_task`	Poll async video task status and download result

These tools require a Pro license and a MINIMAX_API_KEY. Get your key from the MiniMax Interface Key page, then add it to your MCP config — see Environment Variables.

Also ensure your MiniMax account has sufficient balance by topping up at MiniMax Balance before using media generation.

Extended frame extraction

The VIDEO_MCP_MAX_FRAMES setting controls how many keyframes summarize_video and similar tools extract from a video before sending them to the AI:

Tier	Default	Behaviour
Free	50 frames	Clamped between 5 and 50
Pro	100 frames	Default 100; set `0` for uncapped

In both tiers, frame sets are automatically trimmed to fit the active provider’s size or image-count limit (for example: 12 MB for GLM, 10 MB for Qwen/MiMo, 80 MB for Kimi’s frame-fallback path, and 100 images / 100 MB for MiniMax-M3).

Multi-platform video downloads

All users can analyze YouTube videos and direct http(s):// URLs. Pro users unlock downloads from additional platforms:

Vimeo
TikTok
Bilibili
And many other platforms supported by yt-dlp

Free users will see an error message when they pass a non-YouTube platform URL.

Higher resolution frame extraction

Pro tier extracts frames at higher resolution, which improves accuracy for detail-sensitive tasks like reading on-screen text, identifying small objects, or distinguishing similar faces.

Platform video examples

YouTube (works for all users):

“Summarize this YouTube video: https://www.youtube.com/watch?v=abc123”

Vimeo, TikTok, Bilibili (Pro required):

“Transcribe the audio from https://vimeo.com/123456789” “Analyze this TikTok: https://www.tiktok.com/@user/video/123”

Purchase License key

LAUNCH PROMO

For a limited time, you can get a PRO license for just $10. The license is valid for 1 year.

Get license from Gumroad.

The license unlocks all media generation tools — see Model Pricing for what the underlying MiniMax models cost.

Setup your license key

After purchasing, add your license key to the env block of your MCP client configuration file (e.g. .vscode/mcp.json, .mcp.json, or your editor’s MCP settings):

{
  "servers": {
    "videoMcp": {
      "type": "stdio",
      "command": "video-context-mcp",
      "env": {
        "VIDEO_MCP_LICENSE_KEY": "your-license-key-here"
      }
    }
  }
}

You can combine the license key with your other environment variables:

{
  "servers": {
    "videoMcp": {
      "type": "stdio",
      "command": "video-context-mcp",
      "env": {
        "VIDEO_MCP_LICENSE_KEY": "your-license-key-here",
        "GEMINI_API_KEY": "your-gemini-key",
        "AWS_S3_BUCKET": "my-video-analysis-2026"
      }
    }
  }
}