Hi all, We made several updates to the SWE-rebench leaderboard: added new models, refreshed recent results, and reworked the leaderboard UI to make results easier to read, compare, and understand. New Models: Claude Opus 4.8 xhigh: 56.5% — 2.48M tokens GLM-5.2: 51.1% — 2.62M tokens Gemini 3.5 Flash:
