List your company

Web Navigation on Mind2Web Live (test)

52.8Task Completion RateGPT-4o
21.0829.31537.5545.785Feb 17, 2025

Evaluation results

MethodLinks
GPT-4oModel Category=API-based Models2025.0252.858.544.625.3
GPT-4oModel Category=API-based Models2025.0250.456.444.222.1
Qwen2-72B-Instr.Model Category=Open-source Instructed Models2025.0240.9——15.4
Qwen2-72B-Instr.Model Category=Open-source Instructed Models2025.0240.9——15.4
Explorer-7BModel Category=Supervised Fine-Tuning2025.0240.245.334.919.3
Explorer-4BModel Category=Supervised Fine-Tuning2025.0239.44431.318.1
Explorer-7BModel Category=Supervised Fine-Tuning2025.0236.94232.716.4
Explorer-4BModel Category=Supervised Fine-Tuning2025.0236.741.630.816.4
GPT-3.5Model Category=API-based Models2025.0236.5——15.4
GPT-3.5Model Category=API-based Models2025.0236.5——15.4
Qwen2-VL-7BModel Category=Open-source Instructed Models2025.0235.440.234.914.5
Qwen2-VL-7BModel Category=Open-source Instructed Models2025.0233.337.931.712.5
Mistral-7B-Instr.Model Category=Open-source Instructed Models2025.0229.532.824.19.6
Mistral-7B-Instr.Model Category=Open-source Instructed Models2025.0228.6332511.5
Phi-3.5VModel Category=Open-source Instructed Models2025.0223.528.520.52.4
Phi-3.5VModel Category=Open-source Instructed Models2025.0222.32721.21.9