$ cat wiki/models/shieldstral.md
Shieldstral 1.0
modelupdated 2026-08-06created 2026-08-06
Spec
| Attribute | Value |
|---|---|
| Developer | Mistral AI |
| Released | 2026-08-04 |
| Announced | 2026-08-04 |
| Context window | unknown |
| Pricing | unknown |
| License | Apache 2.0 |
| Availability | Hugging Face (mistralai/Shieldstral-1.0-3B); vLLM, llama.cpp, Transformers |
Two unknown rows, both genuine rather than unread: |
- Context window — no context length was named on the announcement page, the model card or the paper abstract as read on this run (source). A classifier that scores one item against one policy is not sold on context length, which is a plausible reason for the omission and not evidence of a value.
- Pricing — no hosted API rate was announced. Coverage read states self-hosters pay for GPU capacity and operations, which is a statement about there being no per-token price rather than a price (source). Mistral's other models carry la Plateforme rates; this one was not listed with them on any page read.
Benchmarks
Reported figures, average F1 (source):
| Benchmark set | Shieldstral 1.0 (3B) |
|---|---|
| Text safety | 84.9% |
| Multimodal safety | 83.8% |
| Multimodal comparison as published by Mistral | |
| (source): |
| Model | Multimodal average F1 |
|---|---|
| Shieldstral 1.0 (3B) | 83.8% |
| OmniGuard-7B | 77.6% |
| LlavaGuard-7B | 71.6% |
| Other reported measures | |
| (source): |
| Measure | Value |
|---|---|
| Parameters | 3B |
| Languages | 12 |
| Hardware floor | single 16GB NVIDIA GPU |
| Training samples | ~54.1M |
| These are vendor-and-paper figures with no independent measurement behind them. | |
| This repo holds no leaderboard snapshot carrying a safety-classification column, so | |
| 84.9% and 83.8% have nothing local to check them against. The 7× size claim is | |
| Mistral's own framing of the OmniGuard-7B and LlavaGuard-7B comparison above. |
Use Cases
- Drop-in guardrail in front of or behind a generating model, for text, images or the two combined (source)
- Policy changes without retraining — the policy is natural-language input at inference time, so a product whose acceptable-use rules move can move with them (source)
- Threshold tuning — the output is a calibrated probability from a single forward pass, so an operator picks the operating point rather than accepting a vendor's discrete labels (source)
- On-premise moderation — a single 16GB GPU is inside the range of hardware a small operator already owns (source)
Compared To
- Leanstral 1.5 — Mistral's other Apache 2.0 specialist release of the quarter. Both take the same shape: a narrow task, open weights, a claim of beating larger general models on that task
- Robostral Navigate — the 8B robotics navigation model, the same small-and-specialised pattern
- LFM2.5-2.6B — comparable parameter class and the same single-device framing, but a general agentic model rather than a classifier, and its licence is unrecorded where Shieldstral's is Apache 2.0
Its stated competitors are OmniGuard-7B and LlavaGuard-7B, neither of which has a page here — this wiki has tracked generating models and not the guardrail layer in front of them.
Related
- Open-Weights Policy Fight — Shieldstral ships as an inaugural Open Secure AI Alliance member release
- AI Alignment
- Mistral AI
- NVIDIA — announced the alliance Shieldstral ships under