Practical Model Menu Setup in Newelle (with Ollama)

Practical Model Menu Setup in Newelle (with Ollama)

Body

Building on the successful first local test run, the next milestone was creating a practical model menu inside Newelle.

To streamline daily use, Ollama models were grouped into categories:

• General Use → qwen2.5:7b-instruct (balanced reasoning + cost-efficient)
• Coding / Development → codellama:13b-instruct and codellama:7b-instruct (programming-focused outputs)
• Fast / Lightweight → tinyllama:1.1b (quick responses for utility prompts)

This menu allows switching between models on demand, based on the task, without reconfiguring endpoints.

➡️ This step makes Ollama integration not just functional, but usable day-to-day within the JARVIS AI Assistant workflow.