AlibabaQwen Guard
Qwen3Guard-Gen-0.6B
Qwen3Guard-Gen treats moderation as a generative instruction task. It is trained from Qwen3 and supports 119 languages and dialects.
Practical use
Practical guidance. For high-volume screening, measure both false alarms and missed cases. Keep the review threshold tied to your content rules rather than treating a model label as a complete policy decision.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | Qwen/Qwen3Guard-Gen-0.6B |
| Training family | The Guard series uses 1.19 million labelled prompts and responses.Qwen3Guard-Gen-0.6B model card |
| Interface | Generative safety classification.Qwen3Guard-Gen-0.6B model card |
| Licence | Apache 2.0.Qwen3Guard-Gen-0.6B model card |
Related models
- Qwen3Guard-Gen-4B
Alibaba · Qwen Guard
Four-billion-parameter generative moderation model trained from Qwen3 for 119 languages and dialects.
Launch date not confirmedDownloadable - Qwen3Guard-Gen-8B
Alibaba · Qwen Guard
Eight-billion-parameter Qwen3-based model that treats text moderation as an instruction task.
Launch date not confirmedDownloadable - Qwen3Guard-Stream-0.6B
Alibaba · Qwen Guard
A 600-million-parameter streaming moderation model with token-level classification.
Launch date not confirmedDownloadable