List your company

Agentic Task Completion on AppWorld Challenge (test)

49.88Task Goal Completion (TGC)Claude-Sonnet-4.5
-1.246412.026825.338.5732Apr 17, 2026

Evaluation results

MethodLinks
Claude-Sonnet-4.5CoEvolve=false2026.0449.8832.37
Gemini-2.5-FlashCoEvolve=false2026.0440.0520.14
Qwen3-30B-A3BCoEvolve=true2026.0431.6516.55
DeepSeek-V3.2CoEvolve=false2026.0422.789.35
Qwen3-30B-A3BCoEvolve=false2026.0419.95.76
Qwen3-4B-InstructCoEvolve=true2026.0417.036.47
LLaMA-3.3-70BCoEvolve=false2026.0416.314.32
GPT-4CoEvolve=false2026.0414.69.3
Gemma-3-27BCoEvolve=false2026.049.830.72
Qwen2.5-7B-InstructCoEvolve=true2026.048.392.16
Qwen3-4B-InstructCoEvolve=false2026.047.912.16
GPT-OSS-20BCoEvolve=false2026.045.520
Qwen2.5-7B-InstructCoEvolve=false2026.040.720