A 4.5M-parameter TTS model runs fully offline on an ESP32-S3
ngxson · x · 2026-11-17
Developer ngxson open-sourced tiny-voca-esp32s3: a 4.5M-parameter English TTS model deployed entirely on an ESP32-S3 board — phonemizer, neural network and speaker all run locally, no cloud or computer needed after flashing. Type a sentence, hear it.
- Supports the Waveshare ESP32-S3-AUDIO board (16MB flash, 8MB PSRAM)
- Requires ESP-IDF v6.1; model export uses Python+PyTorch and weights from Hugging Face
- Repo includes firmware, export scripts, benchmarks and a monitor script, Apache-2.0 licensed
More from Embodied
- NVIDIA open-sources Alpamayo2-Edge, a VLA model for autonomous driving — nvidia · 2026-11-17
- AtomDog: a 3D-printed open-source quadruped robot you control from the browser — Stefania_druga · 2026-11-17
- Open-source LiDAR SLAM hits PyPI, no ROS 2 required: pip install lidarslam — rsasaki0109 · 2026-11-17
- Robotics week in review: Path raises $300M, Runway's Praxis-1 tops open leaderboard — lukas_m_ziegler · 2026-11-17
- Apple robotics paper: 91% success in unfamiliar homes using synthetic sim data — FuSheng_0306 · 2026-11-17
- 'Not in SF, Shenzhen, Bengaluru or Zurich? You're NGMI in robotics' sparks debate — retr0jirachi · 2026-11-17