A community-driven hub for LLM evaluation, learning, and building. Contribute datasets, compete in arenas, and help push the frontier of AI benchmarking — together.
Platform
Community-contributed datasets and puzzle challenges to evaluate the true capabilities of today's leading LLMs. Submit, benchmark, and explore results — all in one open arena.
Deep-dive technical sessions unpacking the core concepts behind modern LLMs.
Top LLM papers, summarized weekly
Latest arena rankings & insights
Featured community puzzle of the week
Weekly on Fridays · No spam · Free
contributors building the world's most community-driven LLM benchmark.