PolicyBench - US Exact Match - tanf: leaderboard
The tanf output of PolicyBench US, without tools. Read countries.us.heatmap exact, with the native common per-output denominator and symmetric reference exclusions. Aggregate programStats rows across models are not model scores. Metric: Exact match (%). Source: https://policybench.org
Metric: Exact match (%). Source: policybench.org. Saturation forecast: Estimated already saturated. 52 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 99 | #99 |
| 2 | Claude Sonnet 4.6 | 99 | #173 |
| 3 | GPT-5.5 | 99 | #50 |
| 4 | Claude Haiku 4.5 | 99 | #580 |
| 5 | DeepSeek V4 Pro | 99 | #198 |
| 6 | Claude Opus 4.7 | 99 | #82 |
| 7 | GPT-5.6 Sol | 99 | #33 |
| 8 | GPT-5.4 Mini | 99 | #401 |
| 9 | Claude Opus 4.8 | 99 | #62 |
| 10 | Gemini 3 Flash (Preview) | 99 | #151 |
| 11 | MiniMax-M3 | 99 | #278 |
| 12 | Gemini 3.8 Flash | 99 | #39 |
| 13 | Claude Opus 5 | 99 | #23 |
| 14 | Gemini 3.5 Flash | 99 | #107 |
| 15 | GPT-6 Astra | 99 | #8 |
No result here: #2 GPT-6.2 Astra, #3 Gemini 4 Carbon, #6 GPT-6.2 Sol.
Interactive version: theaggregate.ai/benchmark?slug=policybench-us-exact-match-tanf · How It Works · Data refreshed daily, snapshot 2026-11-17.