Open performance data for the multichain stack.

A small editorial team and a rotating set of contributors publish one benchmark at a time, each shipping with the script that produced it. The goal is to make performance an observable property of crypto infrastructure — like uptime is for SaaS, or p99 is for databases.

Crypto infrastructure runs the world's open financial rails, and almost none of it is benchmarked in public. Every aggregator, bridge and market-data feed quotes its own numbers, on its own terms, with its own regions and its own definitions of “fast”.

OpenChainBench picks one definition at a time, runs the experiment in the open, and publishes the script alongside the result. We don't aim to embarrass anyone; we aim to make performance a thing you can compare on data, not on marketing copy.

Every benchmark is a YAML spec plus a harness. The spec describes what to measure, which providers, which Prometheus queries hold the numbers; the harness runs continuously and pushes metrics. The site re-queries every minute and re-renders. The leader on every page is computed live from the data.

Anyone can submit a benchmark. The tutorial walks through the four steps. New providers, new metrics, new chains — all welcome via pull request.

• Read the back-catalogue

  Every published benchmark.

• Reproduce any number

  Methodology, harnesses, expected sample sizes.

• Contribute on GitHub

  Pull requests for new providers and benchmarks welcome.

• Follow @openchainbench

  New issues, charts, and corrections.

Bug in a number? Open an issue on GitHub. For everything else, the GitHub repo is the place — discussions, PRs and proposals all go through it.