Multimodal Perception on BabyVision
34.51AccuracyGemini-3-Flash
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-3-Flash2026.02 | 34.51 | |
| Kimi-K2.5 (1T)number of parameters=1T2026.02 | 33.25 | |
| Qwen3-VL-235B-A22Bnumber of parameters=235B2026.02 | 18.3 | |
| ZwZ-8Bnumber of parameters=8B2026.02 | 16.75 | |
| MiMo-VL-7B-RLnumber of parameters=7B2026.02 | 16.24 | |
| ZwZ-4Bnumber of parameters=4B2026.02 | 16.24 | |
| ZwZ-7Bnumber of parameters=7B2026.02 | 15.98 | |
| GLM-4.5V (108B)number of parameters=108B2026.02 | 15.72 | |
| MiniCPM-V-4.5 (9B)number of parameters=9B2026.02 | 14.95 | |
| GPT-5.12026.02 | 13.92 | |
| Qwen3-VL-4Bnumber of parameters=4B2026.02 | 13.66 | |
| Qwen2.5-VL-7Bnumber of parameters=7B2026.02 | 12.89 | |
| Qwen3-VL-8Bnumber of parameters=8B2026.02 | 12.89 |