AppWorld-UL (Function Calling) - Infeasible Tasks: leaderboard

Metric: Interactive task goal completion (%; share of 102 tasks made infeasible, where the agent must tell the user whose AppWorld unit tests all pass, including a test that the agent asked the simulated user for every piece of hidden information it needed; function-calling agent over the AppWorld app APIs; GPT-5.5 user simulator answering from a fixed knowledge set). Source: arxiv.org. Saturation forecast: Around 2031. 5 models tracked.

Top models

#ModelScore
1DeepSeek V4 Pro33.3
2Kimi K2.632.4
3Qwen 3.7 Max28.4
4GPT-5.5 (Medium)28.4
5Claude Opus 4.725.5

Interactive version: theaggregate.ai/benchmark?slug=appworld-ul-function-calling-infeasible-tasks · How It Works · Data refreshed daily, snapshot 2026-09-29.