List your company

Real-World Agent on PinchBench

82.3Average ScoreClaude Opus 4.6
62.85267.90172.9577.999Mar 29, 2026Apr 8, 2026Apr 19, 2026Apr 29, 2026May 10, 2026May 20, 2026May 31, 2026

Evaluation results

MethodLinks
Claude Opus 4.62026.0382.387.4
KAT-Coder-V22026.0381.988.7
MiniMax M2.72026.0381.887.1
GPT-5.42026.0381.690.5
GLM-52026.0380.386.4
SkillAdaptorBackbone=GPT-5.22026.0576.6—
Gemini 3.1 Pro2026.0375.986.7
OpenSpaceBackbone=GPT-5.22026.0575.8—
Base ModelBackbone=GPT-5.22026.0574.8—
SkillAdaptorBackbone=GLM-52026.0573.8—
OpenSpaceBackbone=GLM-52026.0572.3—
Base ModelBackbone=GLM-52026.0570.2—
SkillAdaptorBackbone=Kimi-K2.52026.0567.2—
OpenSpaceBackbone=Kimi-K2.52026.0566—
Base ModelBackbone=Kimi-K2.52026.0563.6—