List your company

Web Reasoning on AssistantBench

31.8AccuracySDP
13.80818.47923.1527.821May 12, 2026

Evaluation results

MethodLinks
SDPLLM=GPT-4o2026.0531.83214.992.854.916
ACP-Domain AgentsLLM=GPT-4o2026.0528.3301167.848.515.5
Magentic-OneLLM=GPT-4o2026.0525.325.31169.935.616.9
SPA-CBLLM=GPT-4T2026.0525.227.59.980.742.712.4
SeeAct→CBLLM=GPT-4T2026.0523.426.19.48247.77.1
CB-1SLLM=GPT-4T2026.0522.224.88.367.849.74.2
RALM-1S→CBLLM=GPT-4T2026.0519.5216.181.3357.3
InfogentLLM=GPT-4o2026.0514.520.45.563.919.38.4