Heads-Up No-Limit Hold'em (HUNL) on GTOWizard benchmark
194mbb/hSlumbot (CFR)
Evaluation results
| Method | Links | |
|---|---|---|
| Slumbot (CFR)Training=Millions CPU-hrs, Interpretable=No, Scales with LLM=No2026.05 | 194 | |
| GPT-5.5 (zero-shot)Training=None, Interpretable=Partial, Scales with LLM=Yes2026.05 | 132 | |
| PokerSkillTraining=None, Interpretable=Yes, Scales with LLM=Yes2026.05 | 61 | |
| Loki (rule-based)Training=None, Interpretable=Yes, Scales with LLM=N/A2026.05 | -200 |