A private, secure, and highly responsive desktop AI assistant. Now featuring Vector Memory, Real-Time Translation, and Context-Aware Suggestions. Your data, your hardware, your rules.
All processing happens on your machine via your Ollama instance. Nothing leaves your system. Zero telemetry, maximum privacy.
Instantly translate AI responses into your preferred language using a dedicated local neural network.
Semantic search embeddings allow Cortex to recall relevant context from far deeper in your history.
Intelligent, context-aware conversation bubbles help keep the dialogue flowing smoothly.
Edit system instructions, fine-tune temperature/seed, and fork conversations directly from the new v1.0.0 UI.
Dual-layer storage with SQLite for chats and Vector Embeddings for long-term semantic recall.
Choose the installation method that suits you.
The complete package including source code and latest features.
Download v1.0.0Clone the repo and install dependencies manually via Python.
View SourceCompiled executables for v1.0.0 are available in the Releases tab.
Download .exeBuilt with PySide6 (Qt). Handles theming, chat bubbles, and the new Suggestion UI.
Coordinates the Synthesis Agent, Title Worker, and new Translation/Suggestion pipelines.
Interfaces with Ollama for Chat, Embedding, and Translation models. Stores data in SQLite + Vectors.