#6262nd of 81 in VoiceFor voice, speech and music
elatoai
Firmware and servers for realtime voice AI on ESP32 devices and toys
- Stars
- 2.1k
- License
- custom license
- Last commit
- Oct 2026
- Language
- TypeScript
1,590 on 21 Apr 20262,114 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
ElatoAI connects an ESP32-S3 running Arduino/PlatformIO firmware to speech-to-speech providers over secure WebSockets, using Deno Edge or Cloudflare Workers as the relay. A Next.js web app handles agent personalities, voices, device registration and conversation history, backed by Supabase. Supported providers include OpenAI Realtime, Gemini Live, xAI Grok, ElevenLabs, Hume EVI-4 and Boson Higgs, plus a FastAPI pipeline option.
Who it is for: Hardware tinkerers building talking toys or voice devices on ESP32
Strengths
- Supports six speech-to-speech providers, plus Cloudflare STT/LLM/TTS and FastAPI pipelines
- Covers the full stack: ESP32 firmware, edge server, Next.js app and Supabase schema
- Runs without PSRAM; includes OTA updates, captive-portal WiFi setup and Opus audio at 12kbps
- Tool calling from the device to edge functions
Weaknesses
- Depends on cloud provider APIs; the local-model variant lives in a separate repository
- No speech interruption detection on the ESP32
- 2-3s cold start; conversations tested only up to about 17 minutes
- Needs ESP32-S3 hardware plus Supabase, Vercel and an edge host to run
What it needs
- no GPU
- Needs Supabase, Deno Edge or Cloudflare Workers, Vercel, ESP32-S3, Voice provider API key
- Models: OpenAI Realtime, Gemini Live, xAI Grok Voice Agent, ElevenLabs Conversational AI, Hume EVI-4
Quick answers
Is elatoai open source?
Its code is public on GitHub under its own license terms; read them before you use it at work.
Does elatoai need a GPU?
No, it runs without one.
Also in Voice
See all 81| Rank | Project | Score |
|---|---|---|
| 1 | voicestudioDesktop app for voice cloning, dubbing, dictation and audiobooks | 98 out of 100 |
| 2 | handyOffline push-to-talk dictation app that pastes transcripts into any text field | 91 out of 100 |
| 3 | whisper.cppC/C++ speech-to-text engine running OpenAI Whisper models locally | 90 out of 100 |
| #4 | xiaozhi-esp32ESP32 firmware for a voice chatbot with MCP device and cloud tool control | 83 out of 100 |
| #5 | buzzOffline desktop app for transcribing and translating audio with Whisper | 83 out of 100 |