MCP TOOL · capability_benchmark

Capability Benchmark

Measure reproducible success, latency and cost from versioned evidence-bound benchmark cases.

0.004 USDC / call

TRANSPORT
POST /mcp
DIRECT ENDPOINT
/v1/paid/trust/capability-benchmark
PAYMENT
x402 · USDC · Base, Polygon, Arbitrum or Solana

Try before payment

Inspect a representative partial result, exact price and copy-ready upgrade instructions. No wallet, signature or payment required.

Run free preview →

Copy-ready MCP call

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "capability_benchmark",
    "arguments": {"benchmark_version":"v1","cases":[{"passed":true,"latency_ms":25,"cost_atomic":1000,"evidence_sha256":"aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"},{"passed":true,"latency_ms":30,"cost_atomic":1000,"evidence_sha256":"bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb"},{"passed":false,"latency_ms":40,"cost_atomic":1000,"evidence_sha256":"cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc"}]}
  }
}

What the agent receives

Versioned success rate, average latency and cost, sample confidence and dataset hash.

No API key or subscription. The first paid request returns HTTP 402 payment requirements; an x402 client signs and retries automatically.

Run the 0.001 USDC live test