S3NS opens generative AI inference on PREMI3NS, billed by the hour
Thales and Google Cloud joint venture S3NS makes model inference services available on PREMI3NS, billed at an hourly rate with unlimited tokens.
By GennnAi··1 min read
The offer bundles Model Garden, Model Registry and an online inference service, with open models from Mistral AI and Google, running on Nvidia H100 GPUs with data processed in France. PREMI3NS holds SecNumCloud 3.2 qualification, but these new AI services still await theirs: notification has been sent to ANSSI, with no completion date. S3NS does not publish its prices. Targeted customers include critical operators, the public sector, healthcare and finance.
Source: IT Social, published on November 5, 2026.