We build your AI applications on the IONOS AI Model Hub: open language models, processed in IONOS data centres in Germany. Your inputs are not stored there and not used for training.
The approved alternative to private AI accounts: a chat for all employees with user management and logging. Every request is processed in Germany.
Answers from your own documents, wikis and ticket systems, always with source references. Optionally including your vendors' documentation.
Ticket triage, email and document processing as workflows, connected via n8n. Customer-facing replies only after approval by your team.
| Topic | According to IONOS documentation |
|---|---|
| Location | Inference exclusively in IONOS data centres in Germany, operational access only from the EU, no third-party sub-processors. |
| Data handling | Prompts and answers are not logged, not stored persistently and not used for training. Only metadata such as time, model and token count is recorded. |
| Certifications | Data centres certified to ISO 27001, the AI Model Hub is within the scope of BSI C5. A data processing agreement is part of the contract. |
| Models | Open language models such as Llama, Mistral, Qwen and GPT-OSS, a coding model, embedding and reranking models for knowledge search and image models from Black Forest Labs. |
| Interface | OpenAI-compatible API, so many existing tools can be connected without rework, including n8n. |
| Billing | Usage-based (tokens or images), no upfront investment in hardware. |
Source: IONOS AI Model Hub documentation (docs.ionos.com), October 2026. Models and details may change; we review them with you in the project.
Wiki, file storage, ticket system or vendor documentation are read in regularly and prepared for search.
Employees or customers ask in the chat. The search finds the relevant sections.
The language model formulates the answer only from these sections and names the source. If knowledge is missing, the AI says so instead of making something up.
All components run in a separate environment per customer in the IONOS Cloud, including the vector database for knowledge search.
For knowledge search, ticket assessment and standard texts, open models are strong enough. Where a frontier model is needed, we advise vendor-neutrally and combine.
In a short live demo we show Secure AI Chat, the knowledge assistant and AI ticket triage, remotely or on site.
An IT problem you would like taken off your plate?
30-min IT consultation