Skip to main content

MCP Server Overview

The JsonCut MCP (Model Context Protocol) Server enables AI assistants to interact with the JsonCut API through a standardized protocol. This allows you to generate images and videos using natural language conversations with your favorite AI tools.

What is MCP?​

The Model Context Protocol (MCP) is an open standard that allows AI assistants to connect to external tools and services. It provides a unified way for AI models to:

  • Access external APIs and services
  • Execute tools and functions
  • Retrieve and manipulate data
  • Interact with specialized systems

JsonCut MCP Server​

JsonCut provides a public MCP server that you can use immediately without any setup or hosting:

https://mcp.jsoncut.com/mcp

Key Features​

  • ✅ Public & Ready to Use - No self-hosting required
  • ✅ Streamable HTTP - Real-time communication via HTTP streaming
  • ✅ Multi-tenant - Each user provides their own API key
  • ✅ Secure - API keys are never stored, only forwarded
  • ✅ Smart Caching - 6-hour schema caching for optimal performance
  • ✅ Full API Access - All JsonCut features available

How It Works​

The JsonCut MCP server acts as a bridge between your AI assistant and the JsonCut API:

  1. You configure your AI assistant with the MCP server URL
  2. You provide your JsonCut API key in the configuration
  3. The MCP server forwards your API key to the JsonCut API
  4. All operations are performed using your account and quota
info

The MCP server acts as a bridge and does not store your API key. It only forwards it with each request to the JsonCut API.

Supported AI Assistants​

The JsonCut MCP server works with any MCP-compatible AI assistant. We provide detailed setup guides for:

What You Can Do​

Once configured, your AI assistant can help you:

Image Generation​

  • Create images from natural language descriptions
  • Build complex multi-layer compositions
  • Apply filters, effects, and transformations
  • Generate images with custom styling and text

Video Generation​

  • Create videos from images and clips
  • Add transitions and effects
  • Combine multiple clips with audio
  • Generate videos with custom timing and animations

Content Management​

  • Upload and manage media files (images, videos, audio)
  • Track job status and download results
  • Organize and retrieve generated media
  • Handle file versioning

Workflow Automation​

  • Generate media based on data or context
  • Create media series with consistent styling
  • Automate content generation
  • Build custom media processing pipelines

Example Conversations​

Simple Image​

You: "Create a social media post with the text 'New Product Launch' on a gradient background."

AI: "I'll create that image for you..."
[Creates image job with text layer and gradient background]

AI: "Done! Your image is ready. Job ID: job_abc123"

Complex Composition​

You: "Create a promotional image with:
- Product photo in the center
- Title text at the top
- Price badge in the corner
- Blur effect on the background"

AI: "I'll build this composition step by step..."
[Creates image job with multiple layers and effects]

AI: "Your promotional image is ready for download!"

Video Generation​

You: "Create a 10-second video from these 3 images with fade transitions"

AI: "I'll create the video..."
[Creates video job with clips and transitions]

AI: "Video generated successfully!"

Getting Started​

To start using the MCP server:

  1. Get a JsonCut API Key

    • Sign up at app.jsoncut.com
    • Navigate to Settings → API Keys
    • Create and copy your API key

  2. Choose Your AI Assistant

    • Select from Claude Desktop, Windsurf, or Cursor
    • Follow the specific setup guide

  3. Configure the Connection

    • Add the MCP server URL
    • Provide your API key
    • Restart your AI assistant

  4. Start Creating

    • Ask your AI to create images or videos
    • The AI will use JsonCut automatically

Technical Details​

Transport Protocol​

The MCP server uses Streamable HTTP for communication:

  • Endpoint: https://mcp.jsoncut.com/mcp
  • Protocol: MCP 2024-11-05
  • Authentication: x-api-key header

Performance​

  • Schema Caching: 6 hours TTL
  • Response Time: Less than 100ms for cached schemas
  • Availability: 99.9% uptime SLA

Rate Limiting​

Rate limits are enforced by the JsonCut API based on your subscription plan:

  • Free Plan: 100 requests/day
  • Hobby Plan: 1,000 requests/day
  • Pro Plan: 10,000 requests/day

Check your dashboard for current limits.

Best Practices​

1. Provide Clear Instructions​

Be specific about:

  • Output format (image or video)
  • Dimensions and aspect ratio
  • Layer composition and order
  • Effects and styling requirements

2. Use Descriptive Layer Names​

For complex compositions:

  • ✅ Name layers clearly
  • ✅ Organize layers logically
  • ✅ Use consistent naming conventions

3. Optimize File Uploads​

For media files:

  • Upload files before referencing them
  • Use appropriate file formats
  • Consider file size limits
  • Set appropriate TTL values

4. Monitor Your Usage​

  • Check your dashboard regularly
  • Monitor API quota usage
  • Upgrade plan if needed

Security & Privacy​

API Key Security​

  • ✅ API keys are transmitted securely via HTTPS
  • ✅ Keys are never stored on the MCP server
  • ✅ Each request is authenticated individually
  • ✅ You can revoke keys anytime in the dashboard

Data Privacy​

  • Your media is processed by JsonCut API
  • Generated files are stored temporarily on the server
  • Files are automatically deleted after TTL expires
  • No data is shared with third parties

Support​

Need help with the MCP integration?

Next Steps​

Choose your AI assistant to get started: