Visual Question Answering on Colon-Bench Unprompted
82.5AccuracyGemini 3 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 3 Pro2026.03 | 82.5 | |
| Gemini 3 ProUnprompted=true2026.03 | 82.5 | |
| Gemini 3 Flash2026.03 | 76 | |
| Gemini 3 FlashUnprompted=true2026.03 | 76 | |
| Seed 1.62026.03 | 72 | |
| Seed 1.6Unprompted=true2026.03 | 72 | |
| Gemini 3.1 Flash Lite2026.03 | 67.7 | |
| Gemini 3.1 F. Li.Unprompted=true2026.03 | 67.7 | |
| Qwen3.5 Plus2026.03 | 60.5 | |
| Qwen3.5 PlusUnprompted=true2026.03 | 60.5 | |
| Qwen3.5 397B2026.03 | 60.1 | |
| Qwen3.5 397BUnprompted=true2026.03 | 60.1 | |
| Qwen3-VL 235B2026.03 | 56.6 | |
| Qwen3-VL 235BUnprompted=true2026.03 | 56.6 | |
| GLM-4.6V2026.03 | 53.8 | |
| GLM-4.6VUnprompted=true2026.03 | 53.8 | |
| Molmo 2-8B2026.03 | 53.4 | |
| Molmo 2-8BUnprompted=true2026.03 | 53.4 | |
| Qwen3-VL Plus2026.03 | 50.5 | |
| Gemini 2.5 Flash Lite2026.03 | 46.8 | |
| Gemini 2.5 F. Li.Unprompted=true2026.03 | 46.8 | |
| Seed 1.6 Flash2026.03 | 45.4 | |
| Qwen-VL Max2026.03 | 45.4 | |
| Seed 1.6 FlashUnprompted=true2026.03 | 45.4 | |
| Qwen-VL MaxUnprompted=true2026.03 | 45.4 | |
| Qwen3-VL 32B2026.03 | 44.4 | |
| Qwen3-VL 32BUnprompted=true2026.03 | 44.4 | |
| Qwen3-VL 8B2026.03 | 38.3 | |
| Qwen3-VL 8BUnprompted=true2026.03 | 38.3 |