notes from the team

The Ekorbia blog.

Release notes, LLM model comparisons, and the occasional local AI feature deep-dive.

Fun local LLM comparisons with Gemma, Granite, and Qwen

Ekorbia v0.2's new compare-chat mode runs 2-3 local models against the same prompt in parallel. I tried Gemma 4, Granite 4.1, and Qwen 3.5 on a few fun prompts. Here's what they said.

Read post →

Ekorbia v0.2 — chat groups and comparison-chat mode

Sidebar folders for organising chat history, side-by-side model comparison with a Keep-this winner pick, a model + tokens footer on every assistant reply, and a piece of database-migration plumbing that'll make every future release smoother.

Read post →

Ekorbia v0.1 — first release of a local LLM client

Ekorbia 0.1.0 is out — a native macOS desktop client for local AI models, running entirely against your own Ollama. Multi-tab chat, folder RAG, prompt library, quick overlay, watches, and more. What's in it and what isn't.

Read post →