HelpBench - Data Concerns: leaderboard
HelpBench (2026) asks LLMs real-world digital privacy, safety and security help-seeking questions and scores each answer against an expert-written rubric of factual and delivery criteria, applied by a Gemini 2.5 Pro auto-rater and averaged over five responses. This board is the rubric score in percent on questions about data collection, exposure, loss and sharing; higher is better.
Metric: Data Concerns topic (data collection, exposure, loss and sharing) rubric score (%): points for met positive criteria and avoided negative criteria (factual criteria written per question by privacy, safety and security experts, plus shared delivery criteria) over the maximum points, applied by a Gemini 2.5 Pro auto-rater at temperature 0; chat versions with default parameters, mean of five responses per question; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.3 | 89 | #197 |
| 2 | Gemini 3 Flash | 88 | #164 |
| 3 | GPT-5 Chat | 87 | #444 |
| 4 | GPT-4.1 | 86 | #495 |
| 5 | Claude Opus 4.6 | 85 | #102 |
| 6 | Grok 4 | 84 | #307 |
| 7 | Gemini 3.1 Pro (Preview) | 83 | #94 |
| 8 | Claude Sonnet 4 | 83 | #400 |
| 9 | GLM-5.1 | 83 | #179 |
| 10 | Gemini 2.5 Pro | 82 | #266 |
| 11 | Claude Sonnet 4.6 | 82 | #161 |
| 12 | Qwen 3.6 Plus | 82 | #222 |
| 13 | GLM-4.6 | 82 | #486 |
| 14 | DeepSeek V3.2 (Non-reasoning) | 81 | #406 (DeepSeek V3.2) |
| 15 | GLM-5-Turbo | 80 | #231 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=helpbench-data-concerns · How It Works · Data refreshed daily, snapshot 2026-11-13.