june

Why we included this project

june is a small, practical way to get a voice assistant running on your own hardware, with no cloud dependency. It pairs an Ollama-hosted LLM with a Hugging Face Whisper model for transcribing your microphone and Coqui TTS for spoken replies, and you can mix text and voice freely on both the input and output sides. Because everything runs locally, none of your conversation data leaves the machine, which is the main draw if privacy is a hard requirement. The JSON config lets you swap in a different model, point at a remote Ollama instance, or clone a voice with an XTTS speaker profile, so it is easy to bend to your own setup. For developers who want a working local voice assistant they can customize, it is a solid starting point rather than a heavyweight platform.