Back to home

AN1 Micro-Benchmark

The AN1 Micro-Benchmark is a public, reproducible harness for measuring externally observable latency and cost deltas when calling the AN1 API.

It treats AN1 strictly as a black box, reports only observable metrics, and makes no assumptions about internal mechanisms. This benchmark is intended for engineers, partners, and evaluators who want to independently verify performance characteristics under controlled conditions.

Access the Public Repository

The benchmark harness is fully open source and available on GitHub. You can clone it, run it against your own workloads, and verify results independently.

What's Inside

  • Reproducible harness for externally observable latency and cost measurements
  • Black-box testing approach with no internal mechanism assumptions
  • Controlled conditions for independent performance verification
  • Complete setup instructions and usage examples
  • Safe for engineering review and pilot evaluation

Request Benchmark API Access

To run the benchmark harness, you'll need a temporary API key. Fill out the form below to request access. Our team will review your request and provide credentials within 1-2 business days.

By submitting this form, you agree to our terms of service and privacy policy.

Built with v0