Vals AI Poker Agent — leaderboard

Metric: TrueSkill Rating. Source: www.vals.ai. 17 models tracked.

Top models

#ModelScore
1GPT-5.21131.83
2GPT-51103.18
3Gemini 3 Flash (Preview)1100.21
4DeepSeek V3.2 (Thinking)1090.3
5Grok 4.1 Fast (Reasoning)1079.22
6Gemini 3 Pro (Preview)1078.9
7DeepSeek V3.11058.23
8Claude Sonnet 4.51055.5
9GPT-5.11038.59
10Grok 4 Fast (Reasoning)1034.3
11Claude Opus 4.5 (20251101)1033.38
12Gemini 2.5 Pro1032.6
13GPT-OSS-120B1015.33
14Kimi K2 (Thinking)1011.63
15Qwen 3 Max (Preview)994.51

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