FEAT REQ - Architect and Editor Model selections for Local LLMs!

[ColumbusAI]

Currently, there is no one-size fits all model for local LLMs on consumer hardware. Yet this split approach has proven to deliver in many tests.

Can Roo-Code add this feature where the user can select separate architect models and editors models?

Look at what can be done if you use a reasoning model for architect and then use coding model for editor mode-

https://x.com/victormustar/status/1898001657226506362

Victor.M.on.X-.QwQ-32B.changed.local.AI.coding.forever.We.now.have.SOTA.performance.at.home.Sharing.my.stack.+.tips.https---t.co-dL3pkCfdm5.-.X.mp4