Butter-Bench: Search for Package
Agents · 2025-10-23
Butter-Bench subtask measuring whether a model can navigate from the charging dock to the kitchen and locate delivery packages.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 2.5 Pro | 100.0 |
| Opus 4.1 | 100.0 |
| Gemini Robotics-ER 1.5 | 100.0 |
| Human (Butter-Bench) | 100.0 |
| GPT-5 | 60.0 |
| Grok 4 | 20.0 |
| Llama 4 Maverick | 0.0 |