Computer Vision Evaluation on CV-Bench
85.8Average ScoreRobobrain-2.0-7B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Robobrain-2.0-7BAccess=Open-Source, Parameters=7B2026.02 | 85.8 | — | — | |
| Gemini-2.5-ProAccess=Closed-Source2026.02 | 85.4 | 79.3 | 91.5 | |
| PhyCritic-7BAccess=Open-Source, Parameters=7B2026.02 | 79.7 | 75.5 | 83.9 | |
| GPT-4oAccess=Closed-Source2026.02 | 79.1 | 73.2 | 85 | |
| Qwen2.5-VL-7BAccess=Open-Source, Parameters=7B2026.02 | 78.9 | 75.1 | 82.7 | |
| Eagle-2.5-8BAccess=Open-Source, Parameters=8B2026.02 | 77.9 | 77.5 | 78.3 | |
| Cosmos-R1-7BAccess=Open-Source, Parameters=7B2026.02 | 75.2 | 74.9 | 75.7 | |
| Gemini-2.5-FlashAccess=Closed-Source2026.02 | 74 | 74.4 | 73.6 | |
| Canvas-of-thoughtBase Model=OpenAI GPT-52026.02 | 61.2 | 52.5 | 70 | |
| Canvas-of-thoughtBase Model=Gemini2.5-pro2026.02 | 60.5 | 52.6 | 68.3 | |
| Iterative ReflectionBase Model=OpenAI GPT-52026.02 | 59.2 | 51.8 | 66.7 | |
| Tree-of-thoughtBase Model=OpenAI GPT-52026.02 | 57.1 | 47.4 | 66.7 | |
| Iterative ReflectionBase Model=Gemini2.5-pro2026.02 | 56.4 | 56.1 | 56.7 | |
| Program-of-thoughtBase Model=Gemini2.5-pro2026.02 | 55.9 | 50 | 61.7 | |
| Chain-of-thoughtBase Model=OpenAI GPT-52026.02 | 55.4 | 47.4 | 63.3 | |
| Program-of-thoughtBase Model=OpenAI GPT-52026.02 | 53.8 | 42.5 | 65 | |
| Tree-of-thoughtBase Model=Gemini2.5-pro2026.02 | 53.7 | 47.4 | 60 | |
| Iterative ReflectionBase Model=OpenAI GPT-o32026.02 | 52.1 | 47.4 | 56.7 | |
| Chain-of-thoughtBase Model=Gemini2.5-pro2026.02 | 51.3 | 45.9 | 56.7 | |
| Iterative ReflectionBase Model=Claude-4-Opus2026.02 | 49.9 | 41.6 | 58.3 | |
| Iterative ReflectionBase Model=Seed-1.6-Thinking2026.02 | 48.5 | 45.3 | 51.7 | |
| Iterative ReflectionBase Model=OpenAI GPT-4o2026.02 | 39.6 | 29.3 | 50 |