AppWorld Challenge: leaderboard
Challenge difficulty split of AppWorld: harder multi-app tasks requiring complex API orchestration across 9 simulated everyday applications.
Metric: Task Goal Completion (%). Source: appworld.dev. Source: appworld.dev. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 85.6 | #49 |
| 2 | GPT-5.6 Luna | 73.4 | #68 |
| 3 | Claude Opus 4.6 | 72.4 | #63 |
| 4 | Qwen 3 14B | 67.6 | #537 |
| 5 | GPT-4.1 | 57.6 | #248 |
| 6 | Qwen 2.5 32B | 47.2 | #546 |
| 7 | GPT-4o | 38.9 | #343 |
| 8 | GPT-4.1 Mini | 37.6 | #355 |
| 9 | GPT-4 Turbo | 17.5 | #404 |
| 10 | Llama 3 70B | 7 | #631 |
Interactive version: theaggregate.ai/benchmark?slug=appworld-challenge · How It Works · Data refreshed daily, snapshot 2026-10-20.