Gemini, ChatGPT, Claude or Groq — pick your model, add your key, and start a natural, memory-keeping conversation on your Echo. No code, no setup hassle.
Leo remembers your name, your preferences, and what you talked about before — across sessions. The mic stays open, so you can speak back to back, naturally.
Leo is provider-agnostic. Pick one of four AI engines, choose your model, and add your own API key. Your key is stored encrypted with KMS.
Adding a new engine? A few lines in a single file.
Find "Leo" in the Alexa app and enable it. One published skill, that's all.
Sign in to the web panel, pick your provider and model, and paste your API key. Your key is encrypted instantly.
Say "Alexa, open Leo" and dive into a natural conversation. Leo handles the rest.
Stores your name and preferences across sessions — you never start from scratch.
A fast model, a tight token budget, and a deadline-cut keep every spoken reply on time.
KMS envelope encryption. Your key is used only to answer on your behalf.
The mic stays open; speak back to back without losing context.
Switch between Gemini, OpenAI, Claude and Groq anytime. Provider-agnostic architecture.
No deploy, no terminal. Set it up in minutes — no technical knowledge required.