List your company

Multimodal Agent Task on GAIA

47.17Success Rate (Level 1)HF Agent
5.965216.662627.3638.0574Apr 30, 2025

Evaluation results

MethodLinks
HF AgentController=GPT-4o2025.0447.1731.411.5433.4
SPORT AgentController=Tuned-Qwen2-VL-7B2025.0435.8516.283.8420.61
HF AgentController=GPT-4o-mini2025.0433.9627.913.8426.06
Warm-up AgentController=GPT-4-turbo2025.0430.215.1017.6
T3-AgentController=MAT-MiniCPM-V-8.5B2025.0426.4211.633.8415.15
T3-AgentController=MAT-Qwen2-VL-7B2025.0426.4215.123.8416.97
HF AgentController=Qwen2-VL-7B2025.0416.988.1409.7
HF AgentController=MiniCPM-V-8.5B2025.0413.215.8107.27
HF AgentController=LLaVA-NeXT-8B2025.049.431.1603.64
HF AgentController=InternVL2-8B2025.047.554.6504.85