Skip to content

Local LLM support #151

@dynodevv

Description

@dynodevv

Describe the Feature Request

Google recently released Gemma 4, including 2 really small yet smart models, Gemma 4 E2B and E4B. These models are specifically optimized to run on smartphones quite quickly. I personally have a phone with a mid-range chip (CMF Phone 1) and the E2B version runs great. My idea basically is that it would be nice if we could download these two models in LastChat, and then use them as we would with models from inference providers. Google also made LiteRT which LastChat could use to integrate the models into the app, since it would handle running the model itself, etc.

Usage Scenario

It would be a nice and private alternative to cloud-based APIs, and it would even work without internet, which is awesome!

Alternatives Considered

None

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions