Build a low-cost conversational device with Higgs Realtime and Elato’s open-source stack. An ESP32-S3, a microphone and a speaker bring live, interruptible speech-to-speech interaction to low-power hardware.

Higgs Avatar can now join a LiveKit room and stream lip-synced video of your voice agent's speech while the agent is still talking. Same higgs-avatar model, same API key.

Join Boson AI and U of T’s MScAC community on October 23–24, 2026. Build across languages, create natural conversations, call tools, and bring expressive avatars to life.

Higgs VoiceOver now works inside Adobe Premiere Pro, bringing directable AI voice-over, word-timed subtitles and custom voices to the Premiere timeline. Powered by Higgs TTS 3.

In Hume AI's independent evaluation of eleven voice-cloning models, Higgs TTS 3 placed fourth on speaker similarity and scored highest on accented reference voices.

Vincent Wargnier and Nayl Ahmadulin won the Higgs Realtime Hackathon with Compass, a voice agent that keeps talking while it works in the background.