AI Intelligence & Model Settings

Gemma 4 & Model Engine Settings

Configure Google AI Studio Gemma 4 for complete repository intelligence, or fallback to local Ollama.

Gemma 4 API Key SpaceDeep Repo Analysis

Power automated architecture pattern extraction, beginner safety scoring, and deep repo understanding.

Gemma 4 / Google AI Studio API KeyStored locally in your browser

Works directly with Google AI Studio keys (AIza...) and OpenRouter keys (sk-or-...).

Provider

Model

Automatically run Gemma 4 Deep Architectural Analysis whenever a new repository is indexed

Architecture Patterns

Automatically discovers whether a repo uses Hexagonal, MVC, or Pipeline design.

Beginner Safe Zones

Flags low-risk modules and safe entrypoints suited for first pull requests.

Zero Zero-Day Leak

Key remains on your device or in server memory; never committed or leaked.

Local Ollama Service: Offline / Standby

Endpoint: http://127.0.0.1:11434

Active Local Model: None

Embeddings Model: nomic-embed-text

Installed Models: 0

Run these in your terminal if you prefer zero-cloud local execution:

ollama pull gemma4:e2b # Gemma 4 lightweight local mentor

ollama pull gemma2:2b # General mentor (~1.6 GB)

ollama pull nomic-embed-text # Lightweight vector embeddings (~274 MB)