GoModel author here. Someone asked me about a benchmark for GoModel during my "Show HN" launch here.

A few days ago, I moved it to a separate repository and gave it another run.

The benchmark is self-reproducible, so you can run it yourself and verify the results.

Feel free to contribute your open-source, self-hosted AI gateway here. If it's mature enough - I'd be more than happy to add it.

(also - if you see any unfairness in the benchmark, please let me know)