Mistral's Shieldstral: a 3B open model that moderates by reading your policy
Mistral released Shieldstral 1.0, a 3-billion-parameter open-weights safety classifier (Apache 2.0) that judges text and images against a moderation policy written in plain language at inference time, returning calibrated probability scores instead of fixed categories. Mistral says it matches or beats open guard models up to seven times its size and runs on a single 16 GB GPU.
Why it mattersEvery app hosting user content needs a filter, and most either rent one from a big provider or bolt on rigid category lists. A small, free, policy-adaptive moderator lowers the cost of doing safety properly — and lets each platform define 'safe' in its own words.
Read in the app — free, in 9 languages
Related stories
DeepSeek's cheap workhorse can now see — and on agent tasks that need eyes it says it is close to Anthropic's best2026-08-21Show the robot once — three to twelve seconds — and it gets the job right 59 times out of 100 with no training at all
2026-08-21The model invents its own tasks, builds the rig to test them, then trains on the results — and DeepReinforce gave the weights away
2026-08-20A free model that fits on one graphics card scores 52 — then spends 22,276 thinking tokens drawing a picture
2026-08-18Zhipu did not build a new model. It kept training the old one — and says coding got 50 percent better.
2026-08-17