A benchmark for evaluating LLM × harness performance.
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).