IONOS AI Model Hub & Cloud GPU

Offering overview

Serverless access to open models is the core of the AI Model Hub, which IONOS runs inside its own cloud data centers and exposes through an OpenAI-compatible REST endpoint. The catalog spans text generation (Qwen3.5 and Qwen3.8 variants, Mistral Nemo and Mistral Small, gpt-oss-120b, Llama 3.3 70B), a coding model (Qwen3-Coder-Next), an OCR vision model, several embedding and reranking models, and FLUX image generators. A playground inside the Data Center Designer lets teams try prompts before writing code, and built-in vector search supports retrieval-augmented generation workflows. Beyond the hosted API, the same cloud offers dedicated GPU servers and Cloud GPU virtual machines for teams that want to train or self-serve models. Data centers are located in Germany, France, the United Kingdom, Spain and the United States, which appeals to European organizations that need GDPR-aligned hosting from a domestic provider. Pricing is strictly pay-per-use with published per-token rates: the cheapest chat model, Qwen3.5-9B, is listed at $0.11 per 1M input tokens and $0.17 per 1M output tokens, embeddings start at $0.01 per 1M tokens, and FLUX.1 schnell images cost $0.032 each.

Technical specifications

Containers

Unknown

Kubernetes

Unknown

Root access

Unknown

API access

Unknown

Autoscaling

Unknown

Persistent processes

Unknown

AI capabilities

Model serving

Qwen3.5, Qwen3.8, Mistral Nemo, Mistral Small 24B, gpt-oss-120b, Llama 3.3 70B, Qwen3-Coder-Next, LightOnOCR 2, bge-m3, FLUX.1 schnell, FLUX.2 klein

AI frameworks

Not stated

Inference endpoints

Unknown

Fine-tuning

Unknown

Vector database

Unknown

Model catalog

Unknown

Monitoring

Not stated

Evaluation

Not stated

Security & compliance

Geography

Provider headquarters are separate from offering infrastructure and availability.

Available plans

Pricing can change. Promotional prices are shown separately from standard prices.

Questions & answers

How do developers access IONOS AI Model Hub?

AI Model Hub provides serverless access to open models through an OpenAI-compatible REST endpoint. A playground in the Data Center Designer lets teams test prompts before writing code.

What types of models does IONOS AI Model Hub offer?

The catalog includes text generation, coding, OCR vision, embedding, reranking and FLUX image generation models.

Does IONOS support retrieval-augmented generation?

AI Model Hub includes built-in vector search for retrieval-augmented generation workflows.

Can teams train or host their own models on IONOS?

IONOS offers dedicated GPU servers and Cloud GPU virtual machines for teams that want to train or self-serve models.

Where are IONOS cloud data centers located?

The supplied record lists data centers in Germany, France, the United Kingdom, Spain and the United States.