Ollama

Ollama runs models locally on your machine — no API key, no data leaving your network. Together with contenox’s local SQLite state, this is the fully sovereign deployment option — see AI sovereignty & the EU AI Act.

Local Ollama

Install Ollama from ollama.com, pull a model, then register it:

ollama pull qwen3:8b

contenox backend add ollama --type ollama
contenox config set default-model qwen3:8b
contenox config set default-provider ollama

Ollama Cloud

Get an API key at ollama.com/settings/keys, then:

export OLLAMA_API_KEY=your-key

contenox backend add ollama-cloud --type ollama --url https://ollama.com/api --api-key-env OLLAMA_API_KEY
contenox model list
contenox config set default-model <name-from-list>
contenox config set default-provider ollama

See also

• Configuration reference