LM Evaluation Harness
Self-Hosted AIUnified framework to benchmark language models on many tasks
Replaces OpenAI Evals
OpenAI Evals is a popular proprietary tool in the self-hosted ai space. These open-source apps let you replace OpenAI Evals with software you host and control yourself.
Moving off OpenAI Evals means trading a managed subscription for a service you run yourself. The payoff is ownership: no per-seat pricing, no vendor lock-in, and full control of your data. Start with the option above that matches your setup-difficulty comfort level — every app links to a detailed specs page and head-to-head comparisons.