PolicyBench - US Exact Match - state_refundable_credits: leaderboard
The state_refundable_credits 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: Around February 2027. 52 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5.2 | 91.84 | #1 |
| 2 | GPT-6.1 Sol | 90.82 | #12 |
| 3 | Claude Fable 5.1 | 90.82 | #17 |
| 4 | Claude Opus 5.5 | 89.8 | #7 |
| 5 | GPT-5.6 Sol | 87.76 | #33 |
| 6 | GPT-6 Astra | 87.76 | #8 |
| 7 | GPT-5.5 | 86.73 | #50 |
| 8 | GPT-6 Luna | 86.73 | #105 |
| 9 | Gemini 3.8 Flash | 85.71 | #39 |
| 10 | Grok 4.7 | 85.71 | #48 |
| 11 | GPT-6 Sol | 85.71 | #29 |
| 12 | Grok 4.6 | 85.71 | #47 |
| 13 | DeepSeek V4.1 Flash | 84.69 | #69 |
| 14 | Gemini 4 Argon | 84.69 | #14 |
| 15 | GPT-6.1 Terra | 84.69 | #22 |
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-state-refundable-credits · How It Works · Data refreshed daily, snapshot 2026-11-17.