BigCodeArena: Judging code generations end to end with code executions

BigCodeArena is a human-in-the-loop platform that evaluates AI code generation by executing code in sandboxed environments across multiple languages and frameworks. It enables interactive testing, multi-turn conversations, and community voting to rank models, addressing key evaluation gaps in code generation. The platform has gathered over 14,000 conversations since February 2025.

Source

HuggingFace Blog

Ingested

OCT 07, 2025 · 19:10

Editorial score

4.0 / 5