Skip to main content
Enclave
Local models

Connect local Ollama models in Enclave (fully offline capable)

Enclave supports local Ollama: point the model layer at an Ollama service running on your machine or LAN, and characters reply with local open-source models (Llama, Qwen, etc.). Combined with self-hosting, this is the strongest privacy posture — neither conversations nor models leave your disk.

Last reviewed:

The short answer

Point Enclave's model layer at local Ollama and characters reply with on-device open-source models — fully offline.

Key points at a glance

  • Local inference: models run on your own machine, no external API key needed.
  • Turn off real-world sync and the whole system makes no outbound requests.
  • Conversations and models both stay on your local disk — the strongest privacy posture.

How to connect

First run Ollama on your machine or LAN and pull a model, then in Enclave's model settings choose Ollama / local-compatible endpoint and enter Ollama's address (default http://localhost:11434). After saving, set the local model as the default or assign it per character.

Who it's for

Users who want maximum privacy, fully offline operation, or no API fees. With MIT open source and one-line docker compose self-hosting, Enclave can be a private world that depends on no cloud at all.

Key facts

Integration
Point to a local Ollama endpoint
Billing
Local inference, no API fees
Offline
Runs fully offline
Data flow
Conversations and models never leave your local disk

Ready to try it?

Open it in your browser — no credit card, no install.

Related

Connect other models

Connect local Ollama models in Enclave (fully offline capable) · Enclave