General Multimodal Understanding on BLINK
73.5AccuracyGemini-2.5-Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-2.5-ProModel Category=Closed-source models2026.06 | 73.5 | |
| GPT-5Model Category=Closed-source models2026.06 | 69.5 | |
| Qwen3-VL-8BModel Category=Open-source MLLMs (no visual tool), Instruct variant=true2026.06 | 69.1 | |
| Qwen3-VL-30B-A3B-InstructInference scheme=no-thinking2026.04 | 67.7 | |
| Qwen3-VL-30B-CoTModel Size=30B, Thinking Mode=CoT2026.02 | 65.4 | |
| Qwen2.5-VL-32BModel Category=Open-source MLLMs (no visual tool), Instruct variant=true2026.06 | 63.6 | |
| GPT-4oModel Category=Closed-source models2026.06 | 63.3 | |
| OmniInference scheme=no-thinking2026.04 | 63 | |
| ReMoT-4b-CoTModel Size=4b, Thinking Mode=CoT2026.02 | 62.2 | |
| TACO (Ours-7B)Model Category=Code-tool / visual-agent models2026.06 | 61.6 | |
| InternVL3.5-30B-A3BInference scheme=no-thinking2026.04 | 60.4 | |
| Qwen3-VL-4B-CoTModel Size=4B, Thinking Mode=CoT2026.02 | 59.5 | |
| Gemini-2.5-Pro2026.02 | 59.1 | |
| GPT-4o2026.02 | 59 | |
| PyVision-RL-7BModel Category=Code-tool / visual-agent models2026.06 | 58.7 | |
| CodeV-RL-7BModel Category=Code-tool / visual-agent models2026.06 | 58.2 | |
| GPT-52026.02 | 57.9 | |
| DeepEyes-v2-7BModel Category=Code-tool / visual-agent models2026.06 | 57.8 | |
| DeepEyes-7BModel Category=Code-tool / visual-agent models2026.06 | 57.5 | |
| InternVL2.5-8BModel Size=8B2026.02 | 56.6 | |
| Pixel-Reasoner-7BModel Category=Code-tool / visual-agent models2026.06 | 56.6 | |
| Qwen2.5-VL-7BModel Size=7B2026.02 | 56.4 | |
| Qwen2.5-VL-7BModel Category=Open-source MLLMs (no visual tool), Instruct variant=true2026.06 | 56.4 | |
| Thyme-7BModel Category=Code-tool / visual-agent models2026.06 | 56.1 | |
| InternVL3-8BModel Category=Open-source MLLMs (no visual tool)2026.06 | 55.5 | |
| Mini-o3-7BModel Category=Code-tool / visual-agent models2026.06 | 55.2 | |
| MathCoder-VL-8BModel Category=Code-tool / visual-agent models2026.06 | 51.7 | |
| LLaVA-Next-7BModel Size=7B2026.02 | 51.3 | |
| LLaVA-OneVision-7BModel Category=Open-source MLLMs (no visual tool)2026.06 | 48.2 |