AgentEscapeBench - Difficulty 15: leaderboard

Metric: Success rate (%; share of the 60 difficulty-15 instances (dependency DAGs of 15 tool and item nodes) whose deterministic final flag the agent submits; single run per instance, official APIs at temperature 0.7, multi-turn tool calling with incremental disclosure of hidden nodes). Source: arxiv.org. Saturation forecast: Around December 2026. 16 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview)83.3
2Claude Opus 4.680
3GPT-5.470
4Kimi K2.570
5Gemini 3 Flash (Preview)50
6GPT-540
7Seed 2.0 Pro31.7
8MiniMax-M228.3
9Qwen 3 235B A22B16.7
10Qwen 3 32B10
11Qwen 3 Next 80B A3B10
12Qwen 3 14B6.7
13Qwen 3 8B0

Interactive version: theaggregate.ai/benchmark?slug=agentescapebench-difficulty-15 · How It Works · Data refreshed daily, snapshot 2026-09-26.