EmbodiedBench ALFRED — leaderboard
Metric: Avg Score (%). Source: embodiedbench.github.io. 24 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.7 Sonnet | 67.7 |
| 2 | Claude 3.5 Sonnet | 64 |
| 3 | Gemini 1.5 Pro | 62.3 |
| 4 | GPT-4o | 56.3 |
| 5 | Gemini 2.0 Flash | 52.3 |
| 6 | Gemini 1.5 Flash | 39.3 |
| 7 | InternVL3-78B | 39 |
| 8 | InternVL3-38B | 38 |
| 9 | InternVL2.5-78B | 37.7 |
| 10 | Gemma 3 27B (IT) | 37 |
| 11 | Gemma 3 12B (IT) | 25.7 |
| 12 | GPT-4o Mini | 24 |
| 13 | InternVL3-8B | 10.3 |
Interactive version: theaggregate.ai/benchmark?slug=embodiedbench-alfred · How the rankings work · Data refreshed daily, snapshot 2026-07-22.