Safety
The model takes moderation policy as a plain-language question and returns a calibrated score. Text and images — one interface. Read the ful…
Mistral AI has released Shieldstral, a 3‑billion‑parameter open‑weight model for content safety that can run locally on device. It interprets plain‑language moderation queries and returns calibrated s
Mistral AI has released Shieldstral, a 3‑billion‑parameter open‑weight model for content safety that can run locally on device. It interprets plain‑language moderation queries and returns calibrated safety scores for both text and images through a single interface. The announcement was made via a Twitter post that received substantial engagement (over 100 thousands of impressions).
Source: Mistral AI (X) | 2026-08-04