RLAMA can't connect to Ollama that runs locally

[textpain]

Hi! Really excited to use this.
However, I ran into some issues. Here's what I've done:

1. curl -fsSL https://ollama.com/install.sh | sh — installed fine
2. ollama run llama3.2:3b — pulled fine, I can talk to the model. If I quit with Ctrl + C, when I try ollama run llama3.2:3b again, it works fine.
3. curl -fsSL https://raw.githubusercontent.com/dontizi/rlama/main/install.sh | sh — installed fine, but complained about Ollama not running
4. ollama serve in a separate terminal — started serving
5. Went to http://localhost:11434 in my browser: It says "Ollama is running"
6. rlama install-dependencies — installed fine, said ✅ All Python dependencies installed successfully in virtual environment!, but again complained about Ollama not running.
7. rlama wizard — hangs for solid ten seconds, then complains that it can't connect to Ollama. Meanwhile, Ollama is running and is accessible. Ollama request log is empty. If I run ollama directly, requests are being shown, so Ollama works and logs requests correctly.
8. rlama --host localhost --port 11434 wizard, rlama --host 127.0.0.1 --port 11434 wizard — same error.

What am I doing wrong? RLAMA version 0.1.39. I'm using Ubuntu 24.

[ShubhamTiwary914]

had the same issue, found that a OLLAMA_HOST env was set to 'http://0.0.0.0:11434' but rlama still coundn't find the ollama serving at localhost:11434

i set that to 'http://127.0.01:11434', and it just seemed to work