Multimodal Reasoning on MMEvalPro
82.8AccuracyThinkLite-VL (+ Ours)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ThinkLite-VL (+ Ours)Model=ThinkLite-VL, Method=Self-Reflection2025.12 | 82.8 | 4.8 | |
| ThinkLite-VL (Vanilla)Model=ThinkLite-VL, Method=Vanilla2025.12 | 78.7 | 13.4 | |
| LEADBase Model=VL-Rethinker-7B, Decoding Strategy=LEAD2026.03 | 75.7 | — | |
| LEADBase Model=VL-Cogito-7B, Decoding Strategy=LEAD2026.03 | 75.6 | — | |
| LEADBase Model=Vision-R1-7B, Decoding Strategy=LEAD2026.03 | 75.1 | — | |
| LEADBase Model=R1-Onevision-7B, Decoding Strategy=LEAD2026.03 | 73.9 | — | |
| LEADBase Model=OpenVLThinker-7B, Decoding Strategy=LEAD2026.03 | 73.5 | — | |
| VL-Rethinker-7BDecoding Strategy=Standard2026.03 | 73.2 | — | |
| VL-Cogito-7BDecoding Strategy=Standard2026.03 | 73 | — | |
| Vision-R1-7BDecoding Strategy=Standard2026.03 | 72.2 | — | |
| OpenVLThinker-7BDecoding Strategy=Standard2026.03 | 71.5 | — | |
| MemVRBase Model=R1-Onevision-7B, Decoding Strategy=MemVR2026.03 | 71.3 | — | |
| SIDBase Model=R1-Onevision-7B, Decoding Strategy=SID2026.03 | 71 | — | |
| VCDBase Model=R1-Onevision-7B, Decoding Strategy=VCD2026.03 | 69.8 | — | |
| R1-Onevision-7BDecoding Strategy=Standard2026.03 | 69.4 | — |