List your company

Code Agent on SWE-Bench Verified

0.809ScoreClaude Opus 4.5
0.67380.70890.7440.7791Jan 6, 2026Feb 4, 2026Mar 5, 2026Apr 3, 2026May 2, 2026May 31, 2026Jun 30, 2026

Evaluation results

MethodLinks
Claude Opus 4.5Framework=OpenHands2026.020.809
Claude-Opus-4.52026.060.809
GPT-5.2 (xhigh)Framework=OpenHands2026.020.8
GPT-5.2 High2026.060.8
Gemini-3-Flash High2026.060.78
GLM-5Framework=OpenHands2026.020.778
Claude Sonnet 4.5Variant=Sonnet 4.52026.010.772
Claude-Sonnet-4.52026.060.772
Kimi K2.5Framework=OpenHands2026.020.768
Seed2.0 Pro2026.060.765
Gemini-3.0 ProVariant=Pro2026.010.762
Gemini 3 ProFramework=OpenHands2026.020.762
Gemini-3-pro High2026.060.762
GPT-5 HighVariant=High2026.010.749
GLM-4.7Framework=OpenHands2026.020.738
Seed2.0 Lite2026.060.735
MiMo-V2-FlashVariant=Flash2026.010.734
DeepSeek-V3.2 ThinkingThinking Mode=true2026.010.731
DeepSeek-V3.2Framework=OpenHands2026.020.731
Kimi-K2 ThinkingThinking Mode=true2026.010.713
GPT-5-mini High2026.060.679