Mistral launches Shieldstral: a 3B open-weight multimodal guardrail that defines safety policies in natural language at inference

Shieldstral

Mistral AI launched Shieldstral on Product Hunt this week—a 3B open-weight multimodal guardrail. You define safety policies in natural language at inference time, and it evaluates text, images, or both with a single token output. It runs locally on one 16GB GPU. The Product Hunt page shows the basics and screenshots, but doesn't disclose latency, accuracy benchmarks, or comparisons with other guardrails like Llama Guard. Treat it as a lightweight, self-hostable compliance component for now; real-world performance needs community benchmarks.

Why it matters: Mistral dropped a 3B multimodal safety guardrail that runs locally and takes plain-language rules—practical for AI app builders. Score held back because the Product Hunt page gives no latency or accuracy numbers, so production readiness is unclear.