Prerequisites
- Ollama installed and running on a PC or server on your network
- At least one model pulled in Ollama (
ollama pull llama3etc.) - Phone and host on the same trusted Wi‑Fi (or VPN)
Steps
- Find your PC's local IP (e.g. 192.168.1.42)
- Ensure Ollama listens on the network interface (check Ollama docs for
OLLAMA_HOST) - In EchoFlow → Settings → add a custom OpenAI-compatible endpoint pointing to your Ollama URL
- Add a model name matching your Ollama model tag
- Start a chat and verify responses
Security
Do not expose Ollama to the public internet without authentication. Use home LAN or a secured VPN. See custom endpoints for broader self-hosted guidance.
Common questions
- Do I need the same Wi‑Fi network?
- Usually yes. Your phone and Ollama host must be on the same LAN unless you use a VPN that routes traffic between them.
- What URL do I use?
- Typically http://YOUR_PC_IP:11434/v1 with an OpenAI-compatible path—confirm in Ollama docs for your version.