PolicyBench - US Exact Match - federal_refundable_credits: leaderboard

The federal_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

#ModelScoreOverall rank
1GPT-6 Astra99#8
2Claude Opus 5.599#7
3GPT-6.1 Sol98#12
4Claude Fable 5.198#17
5Claude Fable 5.298#1
6Claude Sonnet 5.596#21
7GPT-5.6 Sol93#33
8GPT-6 Luna93#105
9GPT-6 Sol93#29
10GPT-6.1 Terra93#22
11Muse Wonder92#31
12Gemini 4 Argon90#14
13Grok 4.690#47
14DeepSeek V4 Pro (0813)90#97
15Kimi K389#59

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-refundable-credits · How It Works · Data refreshed daily, snapshot 2026-11-17.