Sovereign AI – you decide who processes your data

The right model for each task

Choose from a catalogue of providers and models, with price, region and place of processing side by side. Each model is described by what it does – chat, images, embeddings – and by what it is good at, such as coding or reasoning. So you pick by purpose, not by brand.

Your data sets the limit

Every collection, group and provider carries a sovereignty tier – from local (T0) and German servers (T1) to EU servers (T2) and beyond. Before every request the strictest tier involved decides; a request that would break it does not leave your server.

Your assistants keep working

Each assistant has a main and a fallback model. If a provider fails or its quota runs out, the assistant switches by itself – your team notices nothing.

Costs in view

The prices of every offer stand next to the model – you compare before you choose.

Your own tools on the same models

Editors, coding agents and chat apps connect through OpenAI- and Anthropic-compatible gateways – with a key per user and the same rules as everywhere else.