List your company

Visual Question Answering on BC-VL

53.7AccuracyVision-DeepResearch-30B-A3B
8.25220.05131.8543.649Jan 29, 2026

Evaluation results

MethodLinks
Vision-DeepResearch-30B-A3BWorkflow=Agent Workflow2026.0153.7
Claude-3.7-SonnetWorkflow=Agent Workflow2026.0150.4
Gemini-2.5 ProWorkflow=Agent Workflow2026.0149.9
Claude-4-SonnetWorkflow=Agent Workflow2026.0148.6
GPT-5Workflow=Direct Answer2026.0147.2
GPT-5Workflow=Agent Workflow2026.0146.1
Gemini-2.5 FlashWorkflow=Agent Workflow2026.0144.6
Qwen3-VL-30B-A3B-ThinkingWorkflow=Agent Workflow2026.0144.1
Gemini-2.5 ProWorkflow=Direct Answer2026.0143.1
Vision-DeepResearch-8BWorkflow=Agent Workflow2026.0142.6
Qwen3-VL-30B-A3B-InstructWorkflow=Agent Workflow2026.0142.6
Qwen3-VL-8B-InstructWorkflow=Agent Workflow2026.0138.6
Gemini-2.5 FlashWorkflow=Direct Answer2026.0137.1
Qwen3-VL-8B-ThinkingWorkflow=Agent Workflow2026.0137.1
Qwen3-VL-30B-A3B-ThinkingWorkflow=Direct Answer2026.0134.6
Claude-3.7-SonnetWorkflow=Direct Answer2026.0132.3
Qwen3-VL-30B-A3B-InstructWorkflow=Direct Answer2026.0129.6
Claude-4-SonnetWorkflow=Direct Answer2026.0129.3
Webwatcher-32BWorkflow=Multimodal DeepResearch MLLM2026.0126.7
Qwen3-VL-8B-InstructWorkflow=Direct Answer2026.0125.1
Qwen3-VL-8B-ThinkingWorkflow=Direct Answer2026.0125.1
Webwatcher-7BWorkflow=Multimodal DeepResearch MLLM2026.0120.3
Gemini-2.5-flashWorkflow=RAG Workflow2026.0112.1
Qwen-2.5-VL-72BWorkflow=RAG Workflow2026.0110.2
Claude-3.7-SonnetWorkflow=RAG Workflow2026.0110