Multimodal Understanding on MME-RW-en (test)
71.4Overall ScoreSkywork-R1V4
Evaluation Results
| Method | Links | |
|---|---|---|
| Skywork-R1V4Inference Strategy=Agentic, Model Scale=30B2026.02 | 71.4 | |
| ZwZ-8BInference Strategy=Single Forward Pass, Model Scale=8B2026.02 | 69.9 | |
| ZwZ-4BInference Strategy=Single Forward Pass, Model Scale=4B2026.02 | 68.5 | |
| SenseNova-MARS-8BInference Strategy=Agentic, Model Scale=8B2026.02 | 67.9 | |
| ZwZ-7BInference Strategy=Single Forward Pass, Model Scale=7B2026.02 | 66.2 | |
| Qwen3-VL-8BInference Strategy=Single Forward Pass, Model Scale=8B2026.02 | 66 | |
| Qwen3-VL-8B+toolInference Strategy=Tool Use, Model Scale=8B2026.02 | 66 | |
| Mini o3Inference Strategy=Agentic, Model Scale=7B2026.02 | 65.5 | |
| Qwen3-VL-4B+toolInference Strategy=Tool Use, Model Scale=4B2026.02 | 65.4 | |
| DeepEyesV2Inference Strategy=Agentic, Model Scale=7B2026.02 | 64.9 | |
| ThymeInference Strategy=Agentic, Model Scale=7B2026.02 | 64.8 | |
| Pixel-ReasonerInference Strategy=Agentic, Model Scale=7B2026.02 | 64.4 | |
| DeepEyesInference Strategy=Agentic, Model Scale=7B2026.02 | 64.1 | |
| Qwen3-VL-4BInference Strategy=Single Forward Pass, Model Scale=4B2026.02 | 63.5 | |
| Qwen2.5-VL-7BInference Strategy=Single Forward Pass, Model Scale=7B2026.02 | 58.3 |