← Back to Help Center
Integrations
Set up Ollama locally
Use Unimatrix with locally-hosted models via Ollama.
Unimatrix can work alongside a local Ollama installation for local inference while syncing memory to your account (or remaining entirely self-hosted).
Install Ollama and point your client configuration at your local endpoint (default `http://localhost:11434`).
If you self-host the Unimatrix stack via Docker, you can point local classification workers directly at your Ollama instance.
💬
Still need help?
Our developers can assist with custom agent APIs or self-hosting configurations.
Contact Support →