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

#ModelScore
1Qwen 3 14B67.6
2GPT-4.157.6
3Qwen 2.5 32B47.2
4GPT-4o38.9
5GPT-4 Turbo17.5
6Llama 3 70B7

Interactive version: theaggregate.ai/benchmark?slug=appworld-challenge · How the rankings work · Data refreshed daily, snapshot 2026-07-22.