Local Ollama Path

Use this path if you want Omegon working with local inference first. It is the best path for offline evaluation, cost-free experimentation, and environments where external provider setup is a distraction.

What you need

• Ollama installed and running
• At least one model pulled locally
• Omegon installed

Steps

1. Install or start Ollama
2. Pull a model, for example ollama pull qwen3:32b
3. Run omegon in your project directory
4. If needed, use /model and select a local/Ollama-backed model
5. Send this prompt: Read README.md and summarize the project layout.

Success signal

• Omegon responds without asking for an external provider key
• The model/provider metadata indicates a local path
• The agent successfully reads local project files

If it fails

• Confirm ollama list shows at least one installed model
• Confirm the Ollama daemon is reachable from the same machine
• If Omegon still does not detect it, file a Setup problem issue and say you followed the local Ollama path