On June 11, 2026, OVH Groupe announced it is entering into exclusive negotiations to acquire Gladia, a French AI startup specialising in speech-to-text technology. If completed, the deal would fold Gladia’s capabilities directly into OVHcloud and OVHai, making Europe’s largest cloud provider a full-stack player in enterprise voice AI.

This is OVH’s second acquisition as part of its AI Lab strategy, which aims to build the next generation of sovereign generative, agentic, and multimodal AI technologies in Europe.

Who Is Gladia?

Gladia was founded in Paris in 2022 with a focus on making high-quality audio intelligence accessible via API. Their platform handles both real-time and batch speech-to-text transcription across more than 100 languages. As of the announcement, Gladia supports over 300,000 developers and 2,000 enterprise clients.

The appeal is practical. Developers building voice products do not want to manage the complexity of transcription infrastructure. Gladia abstracts that away through a clean API, which is why it found traction quickly in markets where voice workflows are central to the product experience.

Why OVH Is Moving Now

OVH has been building out its AI Lab as part of a broader push to offer European businesses a credible alternative to US-based cloud AI platforms. Data sovereignty concerns and the EU AI Act have created real demand for cloud services that keep data within European borders and under European governance.

Voice AI sits at the intersection of two trends that have accelerated in 2026: the shift to agentic workflows where AI systems take actions rather than just respond, and the demand for enterprise-grade audio intelligence. Customer service, internal reporting, meeting summarisation, and compliance recording are all workflows that depend on reliable transcription as a foundation.

By acquiring Gladia, OVH gains the core building block for a range of voice AI services, from real-time call transcription to audio search and meeting intelligence. The combination positions OVHcloud as a full-stack provider for enterprises that want both compute and voice AI under one sovereign roof.

The Broader Consolidation Picture

This acquisition is not happening in isolation. The voice AI infrastructure market has been consolidating steadily through 2026. Earlier this year, Aircall acquired Vogent. Microsoft released MAI Voice 2 with multilingual support. Anthropic upgraded Claude’s voice capabilities with 18 additional languages and on-the-fly switching.

The pattern is clear: infrastructure-level players are moving to own voice AI end to end rather than relying on third-party APIs. When cloud providers start acquiring transcription startups, it is a signal that voice is becoming table stakes, not a premium add-on.

What This Means for Business

For most businesses, this deal carries a practical message: the cost of enterprise voice AI is going to keep falling, and the reliability is going to keep rising.

When major infrastructure players build transcription and voice intelligence natively into their platforms, it removes a key barrier for businesses that want to deploy voice AI but do not want to stitch together multiple vendors. It also puts competitive pressure on every pure-play voice AI vendor to differentiate on application logic rather than infrastructure.

For businesses running phone-heavy operations, including trades, medical practices, financial advisers, and law firms, the window to act is narrowing in a good way. The technology to answer calls, transcribe conversations, route inquiries, and take action based on what was said is getting cheaper and more accessible every quarter.

The question for most businesses is no longer whether voice AI can handle the job. It is whether they are ready to implement it before their competitors do.

---

If you want to explore how a voice AI employee could work in your business, book a discovery call with Sam.