Enchanted is an iOS and macOS app for chatting with private self-hosted language models through Ollama.
Enchanted solves the problem of accessing large language models on your own terms by providing a ChatGPT-like interface that connects exclusively to privately hosted models such as Llama 2, Mistral, Vicuna, and Starling running on Ollama. The app works by accepting a server endpoint you configure in settings, then routing all conversation through that private instance rather than a cloud service. This approach keeps your data and queries entirely within your control.
The tool suits developers and users who want an unfiltered, secure experience with language models across Apple devices without relying on commercial APIs. It works on macOS, iOS, and visionOS, with features including conversation history stored locally on your device, markdown rendering for formatted responses, image attachments for multimodal prompts, custom system prompts, voice input and text-to-speech output, and offline functionality. The app can be accessed from anywhere on your device through a macOS Spotlight panel and supports custom prompt templates. You will need to run your own Ollama server and, depending on your setup, may need to use a tunneling service to make it accessible to the app.
The project is no longer the primary focus of its maintainers, as indicated by a note in the README stating that Jaz represents the new iteration of the work.
