DPrivBench - Algorithm Level: leaderboard
Algorithm-level differential privacy verification on DPrivBench (2026): for 125 algorithms from the research literature and perturbed versions of them, the model decides whether the stated privacy guarantee holds, no tools, five seeds. F1 of the yes verdict in percent; higher is better.
Metric: F1 score (0-1, times 100) of the yes verdict on the DPrivBench Category 2 algorithm-level instances (125 algorithms from the differential privacy literature with perturbed negative versions); binary yes/no verdict on whether the stated guarantee holds, five seeds, no tools; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-5.5 | High | 82.9 | #54 |
| 2 | Gemini 3.1 Pro (Preview) | 81 | #99 | |
| 3 | GPT-5.4 | High | 79.3 | #135 |
| 4 | Gemini 3 Pro | 74.8 | #145 | |
| 5 | GPT-5 | High | 74.2 | #178 |
| 6 | DeepSeek R1 | 61.4 | #482 | |
| 7 | DeepSeek V3.1 | Non-reasoning | 60.3 | #511 |
| 8 | Gemini 2.5 Flash | 60.2 | #463 | |
| 9 | Claude Sonnet 4.5 | 51.8 | #261 | |
| 10 | Claude Opus 4.5 | 48.7 | #160 | |
| 11 | Qwen 3 30B A3B 2507 Instruct | 46 | #1100 | |
| 12 | Qwen 3 30B A3B 2507 (Thinking) | 40.7 | #764 | |
| 13 | Goedel-Prover-V2-32B | 36.2 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=dprivbench-algorithm-level · How It Works · Data refreshed daily, snapshot 2026-11-21.