Benchmark
Vals Agent Poker Bench
Agentic benchmarks measure how well models carry out tasks that take several steps.
- Models measured
- 17
- Scores recorded
- 17
- Sources
- 1
- IQ weight
- 1.34
Score distribution
- 1,040.8
- Mean
- 1,038.6
- Median
- 57.9
- Std dev
- 994.5
- 10th percentile
- 1,100.2
- 90th percentile
- 890.5
- Min
This benchmark has a weight of 1.34 for chat models, measured over 17 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | GPT 5.2· Vals AI | 1,131.8 |
| 2 | GPT 5· Vals AI | 1,103.2 |
| 3 | Gemini 3 Flash· Vals AI | 1,100.2 |
| 4 | DeepSeek V3.2· Vals AI | 1,090.3 |
| 5 | Grok 4.1 Fast· Vals AI | 1,079.2 |
| 6 | Gemini 3 Pro Preview· Vals AI | 1,078.9 |
| 7 | DeepSeek V3.1· Vals AI | 1,058.2 |
| 8 | Claude Sonnet 4.5· Vals AI | 1,055.5 |
| 9 | GPT 5.1· Vals AI | 1,038.6 |
| 10 | Grok 4 Fast· Vals AI | 1,034.3 |
| 11 | Claude Opus 4.5· Vals AI | 1,033.4 |
| 12 | Gemini 2.5 Pro· Vals AI | 1,032.6 |
| 13 | GPT OSS 120B· Vals AI | 1,015.3 |
| 14 | Kimi K2 Thinking· Vals AI | 1,011.6 |
| 15 | Qwen3 Max Preview· Vals AI | 994.5 |
| 16 | GLM 4.6· Vals AI | 945.8 |
| 17 | Llama 4 Maverick· Vals AI | 890.5 |