Epigeon
Epigeon
Inbox
Primary
Promotions
Social
Docs/Ollama Adapter

On this page

PrerequisitesInstall the CLIInitialize the adapterStart the adapterConnect in EpigeonUseful commandsProduction setup
Back to Docs
Self-hosted AI

Ollama Adapter

Connect your own Ollama server to Epigeon AI through the Epigeon adapter.

AI Settings

Prerequisites

  • Linux server with internet access
  • Node.js 18+
  • Ollama installed and running
curl -sS http://127.0.0.1:11434/api/tags

Install the CLI

npm i -g @epigeon/ollama-adapter-cli

Initialize the adapter

For local use:

epigeon-ollama init

For a hosted tunnel subdomain, open Settings → AI, select Ollama (Epigeon Adapter), click Copy setup command, and run that on your Ollama server instead.

Start the adapter

epigeon-ollama start

Save Adapter token, Public endpoint, and your chosen model name from the output.

Connect in Epigeon

  1. Open Settings → AI
  2. Choose Ollama (Epigeon Adapter)
  3. Enter adapter token, model name, and endpoint
  4. Click Test Provider and save

Useful commands

epigeon-ollama doctor
epigeon-ollama models
epigeon-ollama rotate-token
epigeon-ollama rotate-endpoint
epigeon-ollama status

Production setup

Security

  • Place the adapter behind HTTPS
  • Restrict network access to trusted IPs
  • Rotate tokens regularly

Related Resources

Open AI Settings

Select Ollama (Epigeon Adapter) and test it.

Epigeon AI docs

Learn what Epigeon AI can do once connected.

Install CLI package

Install the user CLI with npm.

PreviousEpigeon AINextCampaigns