General Visual Question Answering on SEEDBench
76.17AccuracyQwen2.5-Omni-3B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-Omni-3BModel Scale=3B2026.06 | 76.17 | |
| AVLM-2BModel Scale=2B2026.06 | 75.1 | |
| Gemma-4-E2B-itModel Scale=4B2026.06 | 64.71 |
Loading the SOTA2 catalog…