Shieldstral-1.0-3B

Mistral AI🇫🇷 France
active
Context window32K tokens

Version History

1.0major

Initial release of Shieldstral-1.0-3B, a 3B-parameter policy-adaptive safety classifier built on Ministral-3-3B-Base-2512 with a Pixtral vision encoder. The model replaces fixed moderation categories with natural-language policy queries evaluated at inference time.

Coverage

model releaseMistral AI

Mistral AI Releases Shieldstral-1.0-3B, a 3B-Parameter Policy-Adaptive Safety Classifier

Mistral AI has released Shieldstral-1.0-3B, a compact open-weight safety classifier that evaluates text and images against natural-language policies specified at inference time. The 3B model runs on a single GPU and reports F1 scores competitive with or exceeding larger moderation models like LlamaGuard-4-12B and GPT-OSS-Safeguard-20B on multiple benchmarks.

3 min read