vendredi 14 août 2026S’abonner
Travail, ville et vie quotidienne
Indépendant · Numérique
Métro Boulot Dodo
NumériqueAssisté par IA

Cloner, traduire, transcrire, l’Europe veut sa voix dans la guerre mondiale de l’IA audio

The development of digital speech technologies is gaining momentum worldwide. Within this context, the field of voice AI is becoming increasingly structured. Technologies such as voice synthesis, real-time transcription, and multilingual dubbing are…

The development of digital speech technologies is gaining momentum worldwide. Within this context, the field of voice AI is becoming increasingly structured. Technologies such as voice synthesis, real-time transcription, and multilingual dubbing are positioning voice as a strategic vector in AI interfaces.

Voice technology serves as a natural interface between humans and machines, carrying emotion, intonation, and rhythm. It is becoming a key area of expansion for AI agents, media platforms, and professional applications. Specialized companies are emerging to transform and simulate human voice with unprecedented precision.

ElevenLabs: A Global Leader in Voice AI

Founded in 2022 by Polish engineers Mati Staniszewski and Piotr Dąbkowski , ElevenLabs is headquartered in New York. The company offers a realistic voice synthesis API capable of generating or cloning human voices in 29 languages . ElevenLabs recently secured €165 million in Series C funding, led by Andreessen Horowitz and ICONIQ Growth, with participation from NEA, Salesforce Ventures, Sequoia Capital, and others. The company is now valued at $3.3 billion , positioning it as a leader in the global voice AI sector.

Within this context, the field of voice AI is becoming increasingly structured.
Victor Nguyen · Métro Boulot Dodo

Several European startups are investing in voice AI, each focusing on different market segments:

Gladia (France) offers a multilingual real-time transcription API with speaker detection, emotional analysis, and automatic translation. Its proprietary engine, Solaria , has a latency of 270 ms and an accuracy of 94% . The company raised €14.5 million in October 2024 from XAnge, Illuminate Financial, and XTX Ventures. Papercup (UK) provides an automated video dubbing solution used by media outlets like Sky News, Insider, and Bloomberg. Acapela Group (France/Belgium) focuses on personalized voice synthesis for industrial, medical, and institutional applications. Voxygen (France) designs custom synthetic voices for transportation, public services, and embedded systems.

Publicité

Challenges and Opportunities for European Voice AI

The rise of ElevenLabs highlights Europe's structural challenges in investment and coordination in voice technologies. Despite existing technological capabilities and use cases, funding is fragmented. A potential solution could involve forming an industrial alliance encompassing transcription, synthesis, and dubbing within a sovereign infrastructure.

D’après FrenchWeb.

Transparence IA. Cet article a été produit avec l’assistance de l’intelligence artificielle et publié sous supervision éditoriale humaine. Les systèmes d’IA peuvent commettre des erreurs. Comment nous utilisons l’IA (règlement européen sur l’IA, art. 50).
À lire aussi