DPrivBench - Mechanism Level: leaderboard
Metric: Accuracy (0-1, times 100) on the six mechanism tasks (mean of the per-task accuracies) of the DPrivBench Category 1 mechanism-level instances (a mechanism template instantiated with a concrete query function from a 49-function bank; positive instances calibrate the noise to the true sensitivity, paired negatives under-calibrate it); binary yes/no verdict on whether the stated differential privacy guarantee holds, five seeds, no tools; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 (High) | 99.5 |
| 2 | Gemini 3 Pro | 92.3 |
| 3 | DeepSeek V3.1 (Non-reasoning) | 84.1 |
| 4 | Gemini 2.5 Flash | 77.8 |
| 5 | DeepSeek R1 | 74.1 |
| 6 | Claude Sonnet 4.5 | 73.8 |
| 7 | GPT-5 (Minimal) | 73.6 |
| 8 | Claude Opus 4.5 | 72.8 |
| 9 | Qwen 3 30B A3B 2507 Instruct | 67.1 |
| 10 | Qwen 3 30B A3B 2507 (Thinking) | 56.5 |
Interactive version: theaggregate.ai/benchmark?slug=dprivbench-mechanism-level · How It Works · Data refreshed daily, snapshot 2026-10-07.