Next-Step-Prediction Style Planning on RoboVQA
64.52Performance ScoreGemini 3 Pro
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Gemini 3 ProScale=-, Mode=Cross-benchmark transfer2026.06 | 64.52 | 58.94 | 1 | |
| Causal PlannerScale=8B, Mode=Cross-benchmark transfer2026.06 | 63.43 | 57.35 | 3 | |
| MiMo-EmbodiedScale=7B, Mode=Cross-benchmark transfer2026.06 | 61.99 | 53.93 | 5 | |
| Seed2.0 ProScale=-, Mode=Cross-benchmark transfer2026.06 | 60.33 | 57.83 | 2 | |
| RynnBrainScale=8B, Mode=Cross-benchmark transfer2026.06 | 60.25 | 54.13 | 4 | |
| RoboBrain-2.5Scale=8B, Mode=Cross-benchmark transfer2026.06 | 59 | 53.55 | 7 | |
| Qwen3-VLScale=8B, Mode=Cross-benchmark transfer2026.06 | 58.55 | 53.04 | 8 | |
| GPT-5.4Scale=-, Mode=Cross-benchmark transfer2026.06 | 58.35 | 52.85 | 9 | |
| Qwen2.5-VLScale=7B, Mode=Cross-benchmark transfer2026.06 | 57.17 | 50.18 | 11 | |
| Cosmos-Reason2Scale=8B, Mode=Cross-benchmark transfer2026.06 | 54.75 | 53.61 | 6 | |
| Kimi K2.5Scale=1T, Mode=Cross-benchmark transfer2026.06 | 53.71 | 50.26 | 10 | |
| RoboBrain-2.0Scale=7B, Mode=Cross-benchmark transfer2026.06 | 46.32 | 37.79 | 16 | |
| Cosmos-Reason1Scale=7B, Mode=Cross-benchmark transfer2026.06 | 43.75 | 44.14 | 12 | |
| GPT-4oScale=-, Mode=Cross-benchmark transfer2026.06 | 34.5 | 43.2 | 13 | |
| Gemini 2.5 ProScale=-, Mode=Cross-benchmark transfer2026.06 | 33.9 | 41.8 | 14 | |
| InternVL3.5Scale=8B, Mode=Cross-benchmark transfer2026.06 | 28.55 | 39.9 | 15 |