Public Benefits Bench v1: leaderboard
Public Benefits Bench v1 reports source-table results for models on helping users navigate SNAP benefits. The evaluated model is measured on text-based public-benefits tasks, and the pinned metric is score (%), reflecting benchmark performance in the mirrored result table. Source: BenchmarkList’s attributed result table.
Metric: Score (%). Source: benchmarklist.com. Saturation forecast: Around July 2027. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 | 71.65 | #20 |
| 2 | Claude Opus 4.8 | 62.11 | #39 |
| 3 | MiniMax-M3 | 60.69 | #151 |
| 4 | Claude Sonnet 4.6 | 58.53 | #94 |
| 5 | Gemini 3.5 Flash | 57.98 | #66 |
| 6 | GLM-5.1 | 57.92 | #109 |
| 7 | DeepSeek V4 Pro | 57.58 | #112 |
| 8 | GPT-5.5 | 57.24 | #33 |
| 9 | Kimi K2.6 | 53.38 | #113 |
| 10 | Gemini 3.1 Pro (Preview) | 50.88 | #64 |
| 11 | Grok 4.3 | 50.07 | #204 |
| 12 | Claude Haiku 4.5 (20251001) | 49.53 | #264 |
| 13 | Grok 4.1 Fast (Reasoning) | 44.32 | #227 (Grok 4.1 Fast) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=public-benefits-bench-v1 · How It Works · Data refreshed daily, snapshot 2026-11-05.