PolicyBench - US Exact Match - federal_income_tax_before_refundable_credits: leaderboard
The federal_income_tax_before_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 | GPT-6 Sol | 90.48 | #29 |
| 2 | GPT-5.6 Sol | 88.1 | #33 |
| 3 | Kimi K3 | 85.71 | #59 |
| 4 | GPT-6 Astra | 84.52 | #8 |
| 5 | Claude Opus 5.5 | 84.52 | #7 |
| 6 | GPT-6.1 Sol | 83.33 | #12 |
| 7 | GPT-6.1 Terra | 83.33 | #22 |
| 8 | Claude Fable 5.2 | 80.95 | #1 |
| 9 | GPT-6 Luna | 79.76 | #105 |
| 10 | Claude Sonnet 5.5 | 79.76 | #21 |
| 11 | Claude Fable 5.1 | 79.76 | #17 |
| 12 | GPT-5.6 Luna | 76.19 | #130 |
| 13 | GLM-5.3 Flash | 75 | #136 |
| 14 | Muse Wonder | 75 | #31 |
| 15 | GPT-5.6 Terra | 72.62 | #68 |
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-federal-income-tax-before-refundable-credits · How It Works · Data refreshed daily, snapshot 2026-11-17.