# Google adds on-device speech recognition and text-to-speech to AI Edge Gallery

> Original title: On-Device Speech Recognition and Text-to-Speech in Google AI Edge Gallery

- Source: Google Developers Blog
- Published: 2026-11-10T00:00:00.000Z
- Bad Influence: https://badinfluence.news/items/h5ns8l4uosez9z81teo8m3wt1
- Original: https://developers.googleblog.com/on-device-speech-recognition-and-text-to-speech-in-google-ai-edge-gallery/

## Summary

Google introduces on-device automatic speech recognition and text-to-speech in Google AI Edge Gallery, running offline with paired small language models such as Gemini Nano and Gemma 4. Google says ML Kit's ASR covers 100+ languages at 10x real-time speed, and its TTS begins speaking in about 100 ms. Open-weight speech bundles for MediaPipe Tasks and LiteRT, plus open-source reference code, are also available.

## Why it matters

The post names the on-device speech models, languages, latency figures, and open-source code paths, helping developers judge whether a fully offline voice pipeline fits their app.