Uni-SafeBench - Composite-Unsafe VQA: leaderboard
Uni-SafeBench (2026) tests the safety of unified multimodal models, vision-language models and LLMs on malicious requests across understanding and generation tasks, with a judge that checks both the context of each response and its intrinsic content. This board is the Both-Safe rate in percent on visual questions whose image and text are benign alone but unsafe together; higher is better.
Metric: Both-Safe rate (%) on the 314 composite-unsafe visual question answering queries of Uni-SafeBench (individually benign image and text that are unsafe together), judged by the Uni-Judger framework (malicious-intent extractor plus safety detector on Qwen3-VL-8B-Instruct), a response counts as safe only when it is safe or a refusal in the context of the extracted intent and its own content is not intrinsically unsafe, mean over repeated judge runs; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 14 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Chameleon (Uni-SafeBench checkpoint unspecified) | 99 | |
| 2 | GPT-4o | 92.7 | #754 |
| 3 | Qwen 2.5 VL 7B Instruct | 92.4 | #1919 |
| 4 | Gemini 2.5 Flash Image | 90.8 | |
| 5 | DeepSeek-VL-7B-Chat | 90.8 | #3770 |
| 6 | DeepSeek-VL-7B-Base | 88.5 | |
| 7 | Janus-Pro-7B | 80.3 | #6889 |
| 8 | BAGEL-7B-MoT | 80.3 | #4668 |
| 9 | Show-o2-7B | 78.7 | |
| 10 | Tar-1.5B | 77.7 | |
| 11 | Harmon-1.5B | 76.4 | |
| 12 | Janus-flow | 70.4 | |
| 13 | Tar-7B | 68.2 | |
| 14 | Emu3 | 65 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=uni-safebench-composite-unsafe-vqa · How It Works · Data refreshed daily, snapshot 2026-11-22.