Nvidia backs Gradium as voice AI startup expands seed round to $100 million

Nvidia backs Gradium as voice AI startup expands seed round to $100 million

The Paris-based voice AI startup's massive seed extension highlights Nvidia's growing influence across every corner of the AI stack, with implications for GPU supply that ripple into crypto mining economics.

Gradium has expanded its seed round to $100 million after adding new investors including Nvidia, giving the Paris based voice AI startup more capital to build real time audio models and expand in the US.

The company said it will use the funding to open a Bay Area office and compete for AI talent closer to the largest US model labs and infrastructure companies. The move puts Gradium nearer to Anthropic, Google, Meta, OpenAI, and Nvidia at a time when voice is becoming one of the main interfaces for AI agents.

Gradium was spun out of Kyutai, the French AI research lab backed by telecom billionaire Xavier Niel. The startup was co founded by Neil Zeghidour, a former researcher at Google Brain, DeepMind, and Facebook.

Advertisement

The company originally launched out of stealth in December with $70 million in seed funding from investors including FirstMark Capital, Eurazeo, DST Global Partners, Eric Schmidt, and Xavier Niel. The latest extension adds roughly $30 million and brings the round to more than $100 million, according to Sifted.

Gradium is building audio models designed for ultra low latency voice generation. In practice, that means AI voices that can respond almost instantly instead of pausing between turns, a problem that still makes many AI agent conversations feel slow or unnatural.

That latency problem has become more important as AI companies push voice agents into customer support, automotive, workplace tools, education, and consumer apps. A voice model that responds quickly enough to feel natural can become an interface layer for agents, not just a text to speech feature.

Gradium says it has already signed customers including Renault, giving it an early commercial foothold in a crowded market. Its competitors include voice AI specialist ElevenLabs, which was valued at $11 billion in February, as well as larger model companies such as Google, OpenAI, and Meta that are also investing heavily in speech and multimodal interfaces.

Nvidia’s participation gives the round extra weight. The chipmaker has become one of the most influential investors in the AI stack, backing companies that could drive future demand for its compute and developer ecosystem. For Gradium, the investment also strengthens its credibility as it tries to recruit in Silicon Valley and sell voice infrastructure to enterprise customers.

The Bay Area expansion is also notable because Gradium is coming from Paris, one of Europe’s strongest AI hubs. France has produced major AI companies and research labs, but Gradium’s move shows that proximity to US talent, customers, and infrastructure partners still matters for frontier AI startups.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
Nvidia backs Gradium as voice AI startup expands seed round to $100 million
Nvidia backs Gradium as voice AI startup expands seed round to $100 million

The Paris-based voice AI startup's massive seed extension highlights Nvidia's growing influence across every corner of the AI stack, with implications for GPU supply that ripple into crypto mining economics.

Share

Add us on Google

Gradium has expanded its seed round to $100 million after adding new investors including Nvidia, giving the Paris based voice AI startup more capital to build real time audio models and expand in the US.

The company said it will use the funding to open a Bay Area office and compete for AI talent closer to the largest US model labs and infrastructure companies. The move puts Gradium nearer to Anthropic, Google, Meta, OpenAI, and Nvidia at a time when voice is becoming one of the main interfaces for AI agents.

Gradium was spun out of Kyutai, the French AI research lab backed by telecom billionaire Xavier Niel. The startup was co founded by Neil Zeghidour, a former researcher at Google Brain, DeepMind, and Facebook.

Advertisement

The company originally launched out of stealth in December with $70 million in seed funding from investors including FirstMark Capital, Eurazeo, DST Global Partners, Eric Schmidt, and Xavier Niel. The latest extension adds roughly $30 million and brings the round to more than $100 million, according to Sifted.

Gradium is building audio models designed for ultra low latency voice generation. In practice, that means AI voices that can respond almost instantly instead of pausing between turns, a problem that still makes many AI agent conversations feel slow or unnatural.

That latency problem has become more important as AI companies push voice agents into customer support, automotive, workplace tools, education, and consumer apps. A voice model that responds quickly enough to feel natural can become an interface layer for agents, not just a text to speech feature.

Gradium says it has already signed customers including Renault, giving it an early commercial foothold in a crowded market. Its competitors include voice AI specialist ElevenLabs, which was valued at $11 billion in February, as well as larger model companies such as Google, OpenAI, and Meta that are also investing heavily in speech and multimodal interfaces.

Nvidia’s participation gives the round extra weight. The chipmaker has become one of the most influential investors in the AI stack, backing companies that could drive future demand for its compute and developer ecosystem. For Gradium, the investment also strengthens its credibility as it tries to recruit in Silicon Valley and sell voice infrastructure to enterprise customers.

The Bay Area expansion is also notable because Gradium is coming from Paris, one of Europe’s strongest AI hubs. France has produced major AI companies and research labs, but Gradium’s move shows that proximity to US talent, customers, and infrastructure partners still matters for frontier AI startups.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.