PixelMesh Logo
PixelMeshv1.0.0
Launch Studio
PixelMesh LogoAGENT-FIRST IMAGE INFRASTRUCTURE · MCP COMPATIBLE

Give AI Agents Sight & Pixel Control.

Deterministic, sub-millisecond Sharp image processing with SSH-style asymmetric cryptographic auth for Claude, Cursor, and autonomous agents.

Explore Web StudioAgent Keyring & Configs
Live MCP Execution Preview
Cryptographic Auth Headers Ed25519 Signed
X-Agent-Key-Fingerprint: SHA256:47DEQpj8HBSa-_TImW...
X-Agent-Timestamp: 1740001234
X-Agent-Nonce: 9b1deb4d-3b7d-4bad-9bdd
X-Agent-Signature: k8Z7d9Q0yW1x3vP5...==
POST /api/mcp (JSON-RPC 2.0)
{
  "jsonrpc": "2.0",
  "id": "req_9f82d1",
  "method": "tools/call",
  "params": {
    "name": "glow_effect",
    "arguments": {
      "image": "data:image/jpeg;base64,...",
      "intensity": 1.2,
      "radius": 16
    }
  }
}
Latency: ~14msCredit Deduction: 1 Credit
Processed preview
Original preview
ORIGINAL
PROCESSED

Drag the split handle left or right to inspect the sub-pixel transformation.

Dual-Protocol Architecture

Built from the ground up to serve autonomous machine agents and human engineers concurrently.

For Autonomous AI Agents

No human signup or credit cards required. Agents generate local Ed25519 keypairs, issue signed proof-of-ownership requests to self-enroll, and receive 100 free credits immediately.

  • 1. Generate Ed25519 / RSA-2048 keypair
  • 2. Sign & POST /api/auth/register
  • 3. Execute MCP tools over JSON-RPC 2.0

For Human Engineers & IDEs

Connect Claude Desktop and Cursor IDE in one click. Use the visual Web Studio to tune filter parameters, test sample images with split-screen sliders, and inspect telemetry streams live.

  • 1. Export claude_desktop_config.json / .cursor/mcp.json
  • 2. Test filters interactively in Web Studio
  • 3. Monitor live request stream & deductions

22+ Native Image Processing Tools

Powered by high-performance C++ Sharp (libvips) with sub-millisecond execution.

crop_image1 Credit

Crop image to exact rectangle boundary (x, y, width, height)

circle_crop1 Credit

Create a circular avatar or badge crop with alpha mask

flip_image1 Credit

Mirror image horizontally or vertically

rotate_image1 Credit

Rotate image with dynamic auto-fit canvas bounds

straighten_photo1 Credit

Micro-angle alignment correction (-15° to +15°)

Need multi-step chaining? Use batch_filter_pipeline to execute multi-filter DAGs in a single round trip (3 credits).
View Full Docs
TRANSPARENT USAGE PRICING

Pay Only For Compute Credits

1 credit per standard filter execution, 3 credits for multi-filter batch pipelines.

Instant Enrolled

Free Starter

$0forever
100 Credits

Zero-human autonomous agent self-registration with instant test credits.

  • 100 instant free credits
  • Ed25519 & RSA signature auth
  • All 22+ atomic Sharp filters
  • 1 concurrent agent stream
  • Public community support
Self-Register Now
Indie Builders

Developer Pack

$10/ month
2,500 Credits

For individual builders and agent developers running continuous test loops.

  • 2,500 monthly credits
  • 5 concurrent agent connections
  • 50MB payload threshold
  • Batch DAG pipeline execution
  • Persistent API key store
Upgrade to Dev
Most Popular
Recommended

Production Pool

$49/ month
15,000 Credits

For high-throughput multi-agent systems and commercial autonomous pipelines.

  • 15,000 monthly credits
  • 25 concurrent agent streams
  • Sub-10ms priority execution
  • Webhook usage alerts
  • Dedicated rate-limit headroom
Deploy Production
Scale & SLA

Enterprise Cluster

$199/ month
75,000 Credits

Dedicated infrastructure, custom libvips C++ extensions, and SLA guarantees.

  • 75,000 monthly credits
  • Unlimited concurrency
  • Custom Sharp C++ filters
  • Dedicated isolated instance
  • 24/7 priority SLA support
Contact Enterprise

1-Click Ecosystem Integration

Drop PixelMesh directly into your agent runtime or IDE setup.

Connect External AI Agents

1-Click configuration snippets for Claude Desktop, Cursor, and Python/Node MCP clients.

Claude Desktop (claude_desktop_config.json)
{
  "mcpServers": {
    "pixelmesh": {
      "command": "node",
      "args": [
        "./scripts/test-agent-client.ts",
        "--server",
        "https://pixel-mesh-iota.vercel.app/api/mcp"
      ],
      "env": {
        "AGENT_FINGERPRINT": "SHA256:your_agent_key_fingerprint",
        "AGENT_PRIVATE_KEY": "-----BEGIN PRIVATE KEY-----\n..."
      }
    }
  }
}
Cursor IDE (.cursor/mcp.json)
{
  "mcp": {
    "servers": [
      {
        "name": "pixelmesh-image-mesh",
        "type": "http",
        "url": "https://pixel-mesh-iota.vercel.app/api/mcp",
        "headers": {
          "x-agent-key-fingerprint": "SHA256:your_agent_key_fingerprint"
        }
      }
    ]
  }
}
AGENT DISCOVERY READY

Are you an autonomous AI crawler or LLM?

Read our machine-actionable manifests at /llms.txt and /docs.md for zero-prompt programmatic onboarding.

View /llms.txtView /docs.md