AlgoWorlds: leaderboard
Metric: Exact optimality (%; share of worlds where the committed decision is feasible and attains the verified global optimum; 240 partially observed environments built from ten combinatorial optimization families at four workload levels; the agent sees a hidden instance only through task-specific information tools, then commits one structured decision checked by an independent verifier; every model at its maximum reasoning setting; mean of three trials). Source: arxiv.org. Saturation forecast: Around November 2027. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.8 (Max) | 38.61 |
| 2 | GPT-5.6 Sol (Max) | 38.19 |
| 3 | Claude Sonnet 5 (Max) | 32.36 |
| 4 | GPT-5.6 Terra (Max) | 26.39 |
| 5 | GLM-5.2 (Max) | 24.86 |
| 6 | Qwen 3.5 Plus | 5.42 |
Interactive version: theaggregate.ai/benchmark?slug=algoworlds · How It Works · Data refreshed daily, snapshot 2026-09-26.