Thanks — great question. Regarding which ESP32 boards you can use with LiveKit: • LiveKit’s ESP32 SDK supports the ESP32‑S3 and ESP32‑P4 devkits. • These boards are optimized for running LiveKit voice AI and real‑time communication features in constrained embedded environments. The ESP32‑P4 also enables H.264 video encoding when you need video streaming. Getting started: • GitHub: https://github.com/livekit/client-sdk-esp32 • API docs: https://livekit.github.io/client-sdk-esp32/ If you tell me which features you plan to use (voice only, voice + video, or data messaging) and which devboard you already have, I can suggest specific examples, required peripherals (microphone, speaker, camera), and steps to get a demo running. Sources: LiveKit SDK for ESP32: bringing voice AI to embedded devices

Based on your requirements, you can select a board that has the peripherals you need. For bidirectional audio applications (like you would need to talk to voice agents), ESP32 S3-based boards are usually the way to go. For development, we usually use the Box-3 from Espressif, but any board that has a microphone and AEC (acoustic echo cancellation) peripherals will work. There are many on the market.

If you let me know your specific requirements, I might be able to recommend a board