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. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 14B | 67.6 |
| 2 | GPT-4.1 | 57.6 |
| 3 | Qwen 2.5 32B | 47.2 |
| 4 | GPT-4o | 38.9 |
| 5 | GPT-4 Turbo | 17.5 |
| 6 | Llama 3 70B | 7 |
Interactive version: theaggregate.ai/benchmark?slug=appworld-challenge · How the rankings work · Data refreshed daily, snapshot 2026-07-22.