List your company

Software Engineering on SWE-Bench Verified (SWE-Agent Metrics)

78.2SWE-Agent ScoreClaude-Opus-4.5
69.8872.0474.276.36Feb 28, 2026

Evaluation results

MethodLinks
Claude-Opus-4.5Size=?, Maximum agent turns=3002026.0278.277.879
Claude-Sonnet-4.5Size=?, Maximum agent turns=3002026.027668.474.6
MiniMax-M2.1Size=230A10, Maximum agent turns=3002026.0274.870.471
GLM-4.7Size=358A32, Maximum agent turns=3002026.0274.270.470.6
Kimi-K2.5Size=1000A32, Maximum agent turns=3002026.0273.270.8—
Qwen3-Coder-NextSize=80A3, Maximum agent turns=3002026.0270.671.171.3
DeepSeek-V3.2Size=671A37, Maximum agent turns=3002026.0270.267.272.6