General Visual Reasoning on MME-RealWorld-Lite
73.06AccuracyGPT4o
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT4oAccess=Close-source2026.02 | 73.06 | |
| S1-VL-32B-RLCategory=Thinking-with-Images Specialist Models, Parameters=32B, Training Stage=RL2026.04 | 67.1 | |
| SAYO-Qwen-8BBase Model=Qwen3-VL-8B2026.02 | 62.85 | |
| S1-VL-32B-SFTCategory=Thinking-with-Images Specialist Models, Parameters=32B, Training Stage=SFT2026.04 | 60.8 | |
| Gemini-3-FlashSize=-2026.06 | 60.34 | |
| Skywork-R1V4-30BCategory=Thinking-with-Images Specialist Models, Parameters=30B2026.04 | 59.3 | |
| PixelEyesSize=8B2026.06 | 59.25 | |
| Gemini 2.5 ProCategory=Proprietary Models2026.04 | 58.3 | |
| SAYO-Qwen-4BBase Model=Qwen3-VL-4B2026.02 | 57.63 | |
| GPT-5Category=Proprietary Models2026.04 | 57.22 | |
| Qwen3-VL-4BParameters=4B2026.02 | 56.8 | |
| Qwen3-VL-235B-A22B-ThinkingCategory=Open-Source Models, Parameters=235B-A22B2026.04 | 56.5 | |
| Intern-S1 (235B+6B)Category=Open-Source Models, Parameters=235B+6B2026.04 | 56.48 | |
| Qwen3-VL-8BParameters=8B2026.02 | 56.23 | |
| ViGoRL2026.02 | 55.34 | |
| Thyme-VL (7B)Category=Thinking-with-Images Specialist Models, Parameters=7B2026.04 | 55.2 | |
| InternVL3.5-38BParameters=38B2026.02 | 54.56 | |
| PixelEyesSize=4B2026.06 | 54.51 | |
| Pixel-ReasonerSize=7B2026.06 | 54.32 | |
| DeepEyesSize=7B2026.06 | 53.53 | |
| NoisyRollout-7BParameters=7B2026.02 | 51.69 | |
| Gemini 2.5 FlashCategory=Proprietary Models2026.04 | 50.2 | |
| ThymeSize=7B2026.06 | 50.13 | |
| SAYO-InternVL-8BBase Model=InternVL3.5-8B2026.02 | 50.03 | |
| Qwen-3-VLSize=8B2026.06 | 49.04 | |
| CVSearchBackbone=LLaVA-OV-7B2026.05 | 48.8 | |
| InternVL3-8BCategory=Open-Source Models, Parameters=8B2026.04 | 48.6 | |
| InternVL3.5-14BParameters=14B2026.02 | 48.57 | |
| OpenVLThinker-7BParameters=7B2026.02 | 48.31 | |
| Qwen3-VL-32B-ThinkingCategory=Open-Source Models, Parameters=32B2026.04 | 47.94 | |
| CVSearchBackbone=Qwen2.5-VL-7B2026.05 | 46.7 | |
| Semantic-back-7BParameters=7B2026.02 | 46.43 | |
| Qwen2.5-VL-32BCategory=Open-Source Models, Parameters=32B2026.04 | 46.2 | |
| CVSearchBackbone=InternVL2.5-8B2026.05 | 46.1 | |
| InternVL3.5-8BParameters=8B2026.02 | 45.34 | |
| InternVL2.5-8B2026.05 | 44.9 | |
| R1-Onevision-7BParameters=7B2026.02 | 44.87 | |
| Qwen-3-VLSize=4B2026.06 | 44.55 | |
| Qwen-2.5-VLSize=7B2026.06 | 44.37 | |
| Qwen2.5-VL-7BCategory=Open-Source Models, Parameters=7B2026.04 | 44.1 | |
| Qwen3-VL-30B-A3BParameters=30B-A3B2026.02 | 43.72 | |
| LLaVA-OV-7B2026.05 | 43.7 | |
| Qwen2.5-VL-7B2026.05 | 42.3 | |
| Mini-o3Size=7B2026.06 | 42.26 | |
| Kimi-VL-16BParameters=16B2026.02 | 42 | |
| Intern-S1-mini (8B)Category=Open-Source Models, Parameters=8B2026.04 | 40.02 | |
| InternVL3.5-30B-A3BParameters=30B-A3B2026.02 | 38.61 |