Benchmarks · economics
Public Benefits Bench v1 (Vals)
Can AI help people navigate SNAP benefits? Scores are the values published by VALS AI. Source benchmark version 1.
9 models · 9 results · updated 2026-06-09 · vals.ai ↗
Leaderboard
Fable 5 max
Heavy line: best score at each price. Thin lines join one model's effort levels. Cost uses each model's median subscription plan, or API price with no plan data. 3 more models under All models in settings.
More from this benchmark
4 published measuresTime per task13 results · lower is better
3 more models under All models in settings.
Score vs any published measure3 measures: sub-scores, splits, counts
Breakdown
darker is better within each column · click a column to sort| Model | |||
|---|---|---|---|
| Fable 5 max | 67.7 | 65.2 | 63.4 |
| Opus 4.8 max | 58.9 | 58.5 | 53.1 |
| MiniMax M3 max | 53.6 | 48.2 | 40.2 |
| Sonnet 4.6 max | 52.4 | 52.4 | 45.6 |
| Gemini 3.5 Flash high | 54.1 | 54.0 | 50.3 |
| GLM 5.1 max | 50.7 | 47.6 | 36.2 |
| GPT-5.5 xhigh | 53.9 | 53.9 | 49.7 |
| Kimi K2.6 max | 43.6 | 46.3 | 34.4 |
| Grok 4.3 high | 44.0 | 46.9 | 38.0 |
All results
9 rows| # | Model | Effort | Score | Subscription $ / task | API $ / task | Time |
|---|---|---|---|---|---|---|
| 1 | Fable 5 | max | 71.7% | $0.12 | $1.05 | 17 min |
| 2 | Opus 4.8 | max | 62.1% | $0.023 | $1.32 | 10.0 min |
| 3 | MiniMax M3 | max | 60.7% | $0.033 | $0.20 | 12 min |
| 4 | Sonnet 4.6 | max | 58.5% | $0.82 | $0.82 | 13 min |
| 5 | Gemini 3.5 Flash | high | 58.0% | $0.78 | $0.78 | 4.6 min |
| 6 | GLM 5.1 | max | 57.9% | $0.088 | $0.34 | 7.3 min |
| 7 | GPT-5.5 | xhigh | 57.2% | $0.056 | $2.02 | 26 min |
| 8 | Kimi K2.6 | max | 53.4% | $0.31 | $0.61 | 9.6 min |
| 9 | Grok 4.3 | high | 50.1% | $1.21 | $1.21 | 7.0 min |