Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-08-05 · high
It runs on one 16GB GPU and marks Mistral's debut in a safety alliance OpenAI, Google, and Anthropic skipped.
Why now: Shieldstral shipped August 4, a week after the Open Secure AI Alliance's July 28 launch, its first product rather than just a founding roster.
Mistral released Shieldstral on August 4, a 3B-parameter safety classifier that scores content against rules written in plain language instead of fixed categories.
That's useful for anyone whose moderation policy changes by jurisdiction or product surface. A new rule is just a rewritten sentence, not a retrained classifier.
Policy evaluation happens at inference time: the model reads the rule and returns a calibrated score from a single token, no fine-tuning required.
The company claims Shieldstral matches open guard models up to seven times its size on text, per the announcement. It also sets a new standard on multimodal moderation across 12 languages.
It runs on one 16GB GPU and handles text and images through the same interface.
Apache 2.0 licensing means teams can inspect, modify, or self-host the model instead of calling a hosted moderation API.
Shieldstral also marks the company's first contribution to the Open Secure AI Alliance, the Nvidia-led group that launched July 28 with 52 partners.
Each link below shares sources, entities, or timing with this story.
NVIDIA released Open Secure AI Alliance / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (NVIDIA released Open Secure AI Alliance); both cover Anthropic, Google, July, Mistral; overlapping topics (anthropic, policy).
NVIDIA released Open Secure AI Alliance / Shared entities / Earlier coverage
Linked by a graph relationship (NVIDIA released Open Secure AI Alliance); both cover Anthropic, Google, July, Nvidia; earlier Anthropic coverage from 2026-07-30.
Linked by a graph relationship (NVIDIA released Open Secure AI Alliance); both cover Anthropic, Google, July, Mistral; earlier Anthropic coverage from 2026-07-27.
NVIDIA released Open Secure AI Alliance / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (NVIDIA released Open Secure AI Alliance); both cover Anthropic, NVIDIA, Open Secure AI Alliance, OpenAI; overlapping topics (alliance, anthropic).
NVIDIA released Open Secure AI Alliance / Shared entities
Linked by a graph relationship (NVIDIA released Open Secure AI Alliance); both cover Anthropic, Google, July, Nvidia.
NVIDIA released Open Secure AI Alliance / Shared entities / Earlier coverage / Tension
Linked by a graph relationship (NVIDIA released Open Secure AI Alliance); both cover Anthropic, Google, NVIDIA, OpenAI; earlier Anthropic coverage from 2026-06-27.
Hugging Face partners with Open Secure AI Alliance / Shared entities / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Hugging Face partners with Open Secure AI Alliance); both cover Anthropic, July, OpenAI; overlapping topics (anthropic, safety).
Anthropic criticizes Qwen / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Anthropic criticizes Qwen); both cover Anthropic, Apache, July; overlapping topics (anthropic, apache, contribution).