aipopdb

9 Nov 2026  ·  6,605 works · 976 news indexed

GR
Lab
Verified 31 October 2026

Gradium

A Paris-based voice AI lab that builds audio language models for real-time voice agents, spun out of the French research lab Kyutai.

Gradium was founded in September 2025 by Neil Zeghidour, a Kyutai founding member who earlier worked on voice models at Google DeepMind, and came out of stealth on 2 December 2025 with a $70 million seed round led by FirstMark Capital and Eurazeo, with backers including Xavier Niel, DST Global Partners and Eric Schmidt. Other coverage names further co-founders (Laurent Mazare, Olivier Teboul, Alexandre Defossez); TechCrunch names only Zeghidour. It launched with English, French, German, Spanish and Portuguese.

Its products are text-to-speech, speech-to-text, live translation, voice cloning and voice design, plus Phonon, an on-device text-to-speech model. In July 2026 it extended its seed round to about $100 million with NVIDIA as a new investor and said it was opening a San Francisco Bay Area office.

Source
2025Founded

Models

2026Phonon