Roboflow Vision Evals - Visual Identification
Multimodal · 2026-07-22
The Roboflow Vision Evals visual-identification task measures exact-match accuracy on 32 image questions that require identifying objects or visual attributes.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Fable 5 | 100.0 |
| Gemini 3.1 Pro Preview | 100.0 |
| Gemini 3.5 Flash | 100.0 |
| Gemini 3.6 Flash | 96.9 |
| Gemini 2.5 Pro | 93.8 |
| Gemini 3 Flash Preview | 93.8 |
| GPT-5.5 | 90.6 |
| Qwen3-VL-235B-A22B-Instruct | 90.6 |
| Muse Spark 1.1 | 87.5 |
| GLM 5V Turbo | 84.4 |
| Qwen3.7-Plus | 84.4 |
| Sonnet 5 | 81.3 |
| Gemini 3.5 Flash-Lite | 81.3 |
| GPT-5.6 Sol | 81.3 |
| Kimi K3 | 81.3 |
| GPT-5.4 Mini | 78.1 |
| GPT-5.6 Luna | 78.1 |
| GPT-5.6 Terra | 78.1 |
| Qwen3.5-27B | 78.1 |
| Opus 4.8 | 75.0 |
| Kimi K2.6 | 62.5 |