About this project

Willow is an open-source, local, and self-hosted voice assistant project presented as a competitive alternative to commercial devices such as Amazon Echo and Google Home. The repository serves as the main project hub for the Willow ecosystem. The README highlights two main points. First, the Willow Inference Server has been released, allowing Willow users to self-host inference for fast language tasks. According to the README, this server supports STT (speech-to-text), TTS (text-to-speech), LLM, and other workloads, and can be used with Willow as well as other applications, including WebRTC. The inference server is maintained in a separate repository (toverainc/willow-inference-server). Second, the README notes that hardware is reaching early users and that GitHub Discussions have been enabled to centralize community conversations. Users are encouraged to introduce themselves and share experiences, with discussions and issues serving as the main support and feedback channels for early adopters. Official documentation is hosted at heywillow.io. The README itself is brief and does not detail installation steps, supported hardware, wake-word behavior, or specific model choices; those details are expected to live in the linked documentation and companion repositories. No performance benchmarks, accuracy figures, or comparisons against other assistants are provided in the README, so none are asserted here.