Managed Ollama hosting

● READY — TESTED END-TO-END

Local LLM runtime (Llama, Mistral, …) — CPU mode with smaller models, your prompts stay with you. We run it for you: one click, your own subdomain, TLS, daily encrypted backups — no Docker, no terminal.

replaces: OpenAI API costs

Launch Ollama →

Free trial — no credit card, cancel monthly.

Facts

How it works

1. Create your account — 30 seconds, no server, nothing to install.
2. Hit launch: we build the isolated environment, issue the TLS certificate and attach the backup stream.
3. Ollama is online at your own subdomain. You keep the keys — credentials are shown once and stored encrypted.

note: First start ~2–5 min per model (download) · CPU mode, smaller models

Honest answers

Is my data locked in?

No. Ollama is open-source standard software — you can export everything and keep running it elsewhere (or yourself) anytime. Cancellation is monthly.

Which plan do I need?

Ollama is resource-hungry, so it needs the Power plan (€39/mo, 4 vCPU · 16 GB RAM). AI apps run on dedicated CPU cores — no GPU, which we say out loud.

Where does it run?

In EU data centers (Germany), operated and supported from Switzerland (Dübendorf ZH).

Related apps