Getleon

Leon is an open-source personal assistant that runs on your own server rather than a third-party cloud, controlled through voice or text commands. Functionality is built from modular "skills" packages, so what Leon can actually do depends on which skills are installed — it's a customizable platform more than a fixed feature set. Who it's for: privacy-conscious users and developers who want an assistant they fully control — self-hosted, with the underlying code auditable — rather than a cloud assistant with an opaque, centrally-hosted backend. Being open source, it particularly suits people comfortable running and maintaining their own server-based tools. Strengths: full data control from self-hosting is a real, structural difference from most commercial voice assistants, and the MIT license means it's genuinely free to use, modify, and redistribute with no vendor lock-in. It supports multiple speech-to-text and text-to-speech backends (Google Cloud, AWS, IBM Watson, and others), so you aren't tied to one provider's voice pipeline. Weaknesses and limitations: self-hosting means you're responsible for setup, maintenance, and keeping the server running — a materially bigger commitment than installing a consumer app, and the actual capability set depends entirely on which community skills exist and are installed. It's maintained primarily by its creator's own time and community contributions, not a company with a support team behind it. Real-world scenario: a developer wants a voice-controlled assistant for home automation or personal task automation, and specifically wants it running on hardware they control rather than sending voice data to a third-party cloud service.