TaxBench — leaderboard
TaxBench evaluates AI models on real-world tax tasks from Rivet's active tax workflows, spanning tax knowledge and judgment, tax calculations, and agentic data-retrieval question answering.
Metric: Mean pass^5 (computed) (self-reported). Source: benchmarklist.com. Status: saturation imminent. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 Pro | 29.27 |
| 2 | GPT-5.4 Pro (xHigh) | 27.03 |
| 3 | GPT-5.5 | 24.43 |
| 4 | Claude Opus 4.6 | 21.37 |
| 5 | Gemini 3.1 Pro (Preview) | 20.1 |
| 6 | Grok 4.1 Fast | 19.63 |
| 7 | GPT-5.2 Pro | 17.77 |
| 8 | Gemini 3.1 Flash | 15.93 |
| 9 | Grok 4.20 (Reasoning) | 15.2 |
| 10 | Claude Opus 4.7 | 14.37 |
| 11 | Grok 4.20 | 12.23 |
| 12 | Claude Sonnet 4.6 | 11.2 |
| 13 | GPT-5.4 | 9.33 |
| 14 | Gemini 2.5 Pro | 9 |
| 15 | Claude Sonnet 4.5 | 8.03 |
Interactive version: theaggregate.ai/benchmark?slug=taxbench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.