ALFWorld: leaderboard
ALFWorld: Measures embodied-agent, navigation, manipulation, or simulated robotics task success.
Metric: ALFWorld normalized score. Source: alfworld.github.io. Source: benchmarklist.com. Saturation forecast: Around June 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 (High) | 1 | #63 (Claude Opus 4.6) |
| 2 | Claude Sonnet 4.6 (High) | 1 | #90 (Claude Sonnet 4.6) |
| 3 | Claude Opus 4.5 (High) | 1 | #83 (Claude Opus 4.5) |
| 4 | GPT-5 (High) | 0.93 | #98 (GPT-5) |
| 5 | Claude Sonnet 4 | 0.92 | #199 |
| 6 | GPT-5.1 (High) | 0.92 | #135 (GPT-5.1) |
| 7 | O3 (Medium) | 0.88 | #124 (O3) |
| 8 | Claude 3.7 Sonnet | 0.83 | #249 |
| 9 | O3 (High) | 0.82 | #124 (O3) |
| 10 | O3 (Low) | 0.7 | #124 (O3) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=alfworld · How It Works · Data refreshed daily, snapshot 2026-10-20.