List your company

Fact-based Visual Question Answering on FVQA

74.2AccuracyVision-DeepResearch-30B-A3B
15.4430.69545.9561.205Jan 29, 2026Feb 15, 2026Mar 4, 2026Mar 21, 2026Apr 7, 2026Apr 24, 2026May 11, 2026

Evaluation results

MethodLinks
Vision-DeepResearch-30B-A3BWorkflow=Agent Workflow2026.0174.2
POINTS-Seeker-8BWorkflow=Agentic Search MLLM2026.0471.2
Qwen3-VL-30B + ODE-RLEvaluation Setting=Visual-Native Agent Harness, Training Strategy=ODE-RL2026.0569.3
MM-DeepResearch-8BWorkflow=Agentic Search MLLM2026.0469.2
GPT-5Workflow=Agent Workflow2026.0169
Claude-4-SonnetWorkflow=Agent Workflow2026.0169
GPT-5Workflow=Agent Workflow2026.0469
Claude-4-SonnetWorkflow=Agent Workflow2026.0469
Claude-4-SonnetEvaluation Setting=Agent Workflow2026.0569
Gemini-2.5 ProWorkflow=Agent Workflow2026.0168.3
Gemini-2.5 FlashWorkflow=Agent Workflow2026.0168
Gemini-2.5 FlashWorkflow=Agent Workflow2026.0468
Claude-3.7-SonnetWorkflow=Agent Workflow2026.0167.3
Claude-3.7-SonnetEvaluation Setting=Agent Workflow2026.0567.3
Skywork-R1V4-30B-A3BWorkflow=Agentic Search MLLM2026.0467.2
SenseNova-MARS-8BWorkflow=Agentic Search MLLM2026.0467.1
Qwen3-VL-30B + ODE-SFTEvaluation Setting=Visual-Native Agent Harness, Training Strategy=ODE-SFT2026.0566.3
Gemini-2.5 ProEvaluation Setting=Agent Workflow2026.0565
Vision-DeepResearch-8BWorkflow=Agent Workflow2026.0164.7
Vision-DeepResearch-8BWorkflow=Agentic Search MLLM2026.0464.7
Qwen3-VL-8B + ODE-RLEvaluation Setting=Visual-Native Agent Harness, Training Strategy=ODE-RL2026.0564.7
Qwen3-VL-30B-A3B-ThinkingWorkflow=Agent Workflow2026.0163
GPT-5Evaluation Setting=Agent Workflow2026.0562
Qwen3-VL-8B + ODE-SFTEvaluation Setting=Visual-Native Agent Harness, Training Strategy=ODE-SFT2026.0562
Gemini-2.5 ProWorkflow=Direct Answer2026.0160.7
Gemini-2.5 ProEvaluation Setting=Direct Reasoning2026.0560.7
DeepEyesV2Workflow=Agentic Search MLLM2026.0460.6
Qwen3-VL-8B-InstructWorkflow=Agent Workflow2026.0158.7
MMSearch-R1-7BWorkflow=Multimodal DeepResearch MLLM2026.0158.4
MMSearch-R1-7BWorkflow=Agentic Search MLLM2026.0458.4
MMSearch-R1-7BEvaluation Setting=Deep Search Agent2026.0558.4
Qwen3-VL-30B-A3B-InstructWorkflow=Agent Workflow2026.0157.7
GPT-5Workflow=Direct Answer2026.0157.3
GPT-5Workflow=Direct Answer2026.0457.3
GPT-5Evaluation Setting=Direct Reasoning2026.0556.3
Gemini-2.5 FlashEvaluation Setting=Agent Workflow2026.0554.7
Qwen3-VL-30BEvaluation Setting=Visual-Native Agent Harness2026.0554.3
Qwen3-VL-8B-ThinkingWorkflow=Agent Workflow2026.0151.3
Qwen3-VL-8B-ThinkingWorkflow=Agent Workflow2026.0451.3
Gemini-2.5 FlashWorkflow=Direct Answer2026.0147.7
Gemini-2.5 FlashWorkflow=Direct Answer2026.0447.7
Gemini-2.5 FlashEvaluation Setting=Direct Reasoning2026.0547.7
Qwen3-VL-8BEvaluation Setting=Agent Workflow2026.0545.3
Qwen3-VL-30BEvaluation Setting=Agent Workflow2026.0545.3
Qwen3-VL-8BEvaluation Setting=Visual-Native Agent Harness2026.0544.7
Claude-3.7-SonnetWorkflow=Direct Answer2026.0136.7
Claude-3.7-SonnetEvaluation Setting=Direct Reasoning2026.0535.7
Claude-4-SonnetWorkflow=Direct Answer2026.0135.3
Claude-4-SonnetWorkflow=Direct Answer2026.0435.3
Claude-4-SonnetEvaluation Setting=Direct Reasoning2026.0535.3
Qwen3-VL-30B-A3B-InstructWorkflow=Direct Answer2026.0134.7
Qwen3-VL-30BEvaluation Setting=Direct Reasoning2026.0534.7
Qwen3-VL-30B-A3B-ThinkingWorkflow=Direct Answer2026.0132.7
MiMoVL 7B-RLRL=MiMoVL, Initial Model=7B-SFT2026.0431.8
GPT-5 NanoRL=N/A, Initial Model=N/A2026.0429.4
Qwen3-VL-8B-InstructWorkflow=Direct Answer2026.0128
Qwen3-VL-8B-InstructWorkflow=Direct Answer2026.0428
Qwen3-VL-8BEvaluation Setting=Direct Reasoning2026.0528
Vero Mi-7BRL=MiMoVL, Initial Model=7B-SFT2026.0427.9
Vero Q25-7BRL=Qwen25VL, Initial Model=7B Inst2026.0426.3
Q3VL 8B-InsRL=N/A, Initial Model=N/A2026.0426
Vero Q3I-8BRL=Qwen3VL, Initial Model=8B Inst2026.0424.6
Q3VL 8B-ThkRL=N/A, Initial Model=N/A2026.0424.2
Qwen3-VL-8B-ThinkingWorkflow=Direct Answer2026.0124
Vero Q3T-8BRL=Qwen3VL, Initial Model=8B Think2026.0422
Q25VL 7B-InsRL=N/A, Initial Model=N/A2026.0421.9
Mo2-O 7BRL=SFT Only, Initial Model=N/A2026.0417.7