Incorrect model type detection for gemma3:12b in KnownNote (Ollama provider)

[Diyago]

Issue: Incorrect model type detection for gemma3:12b in KnownNote (Ollama provider)

KnownNote incorrectly classifies the Ollama model gemma3:12b as unknown instead of identifying it as a chat model.

Expected behavior
When fetching models from the Ollama API (/api/tags), KnownNote should auto-detect gemma3:12b as a chat model, similar to other LLMs.

Actual behavior

gemma3:12b appears in the “Other Models” list with the type unknown.

It does not show up under “Chat Models”, even though it is a full LLM.

Manual editing of the model entry is currently not possible: KnownNote does not allow users to reassign a model type.

Environment

Ollama running locally

Model pulled with: ollama pull gemma3:12b

KnownNote version: [your version]

OS: Windows 10

Steps to reproduce

Install Ollama

Pull gemma3:12b

Open KnownNote → AI Providers → Ollama → Fetch Models

Observe model classification

Proposed improvement

Allow users to manually set or change model type (Chat / Embedding / Completion).

Add explicit mapping for gemma3:* models as chat models.

Or infer model type from metadata returned in Ollama’s /api/generate capabilities.

Impact
Users cannot use gemma3:12b for chat tasks unless manually patched, limiting integration with Ollama models.

[Diyago]

workaround

ollama cp gemma3:12b gemma3-chat:latest

[MrSibe]

Thanks for feedback! I will try to fix this in the next release.