DeepSeek V4.1 Flash Becomes First Open-Weight Model to Exceed 1% on SWE-sweep

KLieret's updated SWE-sweep leaderboard shows DeepSeek V4.1 Flash (max) scoring 1.7%, making it the first open-weight model to exceed 1% on the notoriously difficult bug-hunting benchmark.

2026-11-14 ~ 2026-11-14 · 2 related posts

• DeepSeek 4.1 Flash Max Is First Open-Weight Model to Top 1% on Bug-Finding Benchmark — mariofilhoml · 2026-11-14
• DeepSeek V4.1 Flash is first open-weight model to top 1% on SWE-sweep bug benchmark — klieret · 2026-11-14