ElevenLabs and NVIDIA expand collaboration to bring voice agents to enterprises and governments globally

• ElevenLabs and NVIDIA are expanding their work across models, infrastructure, and sovereign deployments to help enterprises and governments run voice agents at scale.
• NVIDIA Nemotron 3.5 Super VL is now available in ElevenAgents, and ElevenLabs is scaling its NVIDIA infrastructure by adding 15MW of compute based on NVIDIA Blackwell Ultra systems and NVIDIA Vera Rubin platform.
• ElevenLabs is using its voice AI models and NVIDIA infrastructure to power a sovereign deployment of the voice agent for Poland’s national e-Health platform.
• ElevenLabs is also exploring on-device voice AI for robotics and physical AI, combining its multilingual audio models with NVIDIA edge computing platforms to enable more natural human-machine interaction.

NEW YORK, 11 NOVEMBER 2026 — ElevenLabs and NVIDIA have today expanded their partnership, bringing NVIDIA’s models, infrastructure, and secure deployment to ElevenLabs’ conversational agents platform ElevenAgents.

“As voice agents become critical infrastructure, businesses need to be able to operate securely, efficiently, reliably, and with control over their data,” said Mati Staniszewski, co-founder and CEO of ElevenLabs. “By combining our audio models and agents, with NVIDIA’s LLMs, accelerated computing and security stack, we’re giving enterprises and governments the platform they need to deploy voice agents at massive scale. NVIDIA has been a superb partner to us, and we’re excited to keep building together.”

“For decades we adapted to computers — keyboards, screens, menus. Now computers are learning to speak our language,” said Jensen Huang, founder and CEO of NVIDIA. “With NVIDIA Nemotron reasoning inside ElevenAgents and NVIDIA accelerated computing powering large-scale deployments, enterprises and governments can run voice agents securely at national scale — and make talking the fastest way to get things done.”

Nemotron joins the models available in ElevenAgents

Nemotron open models are designed for agentic workloads that require models to reason, use tools and execute complex, multi-step tasks efficiently. Enterprises can combine them with ElevenLabs’ low-latency, multilingual voice models to create systems of models for building voice agents that operate efficiently at scale.

On the ElevenAgents benchmark for realistic customer-support tasks, Nemotron 3.5 Super VL achieved leading end-to-end task-completion accuracy—including strong performance in financial calculations and billing analysis—at 70% lower costs, compared to other leading open models. Nemotron offers the option to turn reasoning off for simpler tasks, placing it among the three fastest models on the ElevenAgents platform.

Nemotron open models are easy to post-train with enterprise data and expertise, enabling organizations to build, own and continuously improve their AI. ElevenLabs customers can also use these post-trained models to power voice agents in ElevenAgents.

More capacity for the largest deployments

To meet growing demand from its largest customers, ElevenLabs is expanding its on-premises infrastructure. The company plans to add 15 MW of next-generation NVIDIA Vera Rubin capacity, on top of its existing NVIDIA Blackwell Ultra systems. The added capacity supports large-scale deployments, including those with strict security and data-residency requirements, and gives ElevenLabs’ research teams access to optimized compute.

NVIDIA and ElevenLabs have worked together since 2022. ElevenLabs’ first text-to-speech model was trained on NVIDIA GPUs, and NVIDIA has been using ElevenLabs’ technology since 2024, to generate voiceovers for 50+ videos across all major NVIDIA events, including translations for Jensen Huang’s keynotes. In 2025, NVIDIA also became a strategic investor in ElevenLabs.

Improving access to healthcare in Poland

ElevenLabs is using NVIDIA technology on sovereign deployments, starting in Poland. ElevenLabs models and NVIDIA infrastructure power the voice agent for Centralna e-Rejestracja, the central appointment system within Poland’s national e-Health platform with EasyCall supporting the delivery of the voice agent solution.

The agent calls patients who only have landlines to remind them of upcoming appointments and help them reschedule. With 40 million appointments a year and no-show rates of 10 to 20 percent, even a small reduction in missed visits can significantly improve access to care. 

The deployment runs on-premises using NVIDIA Confidential Computing, keeping citizen data and audio within government infrastructure and protected during processing, and local deployment reduces latency for real-time public services.

Bringing voice to physical AI

Today, as leading industrial, automotive, and manufacturing companies explore on-device voice AI for robotics and physical AI, ElevenLabs and NVIDIA are working to make it easier for people to talk naturally to the industrial machines, robots, and vehicles they use every day.

By bringing ElevenLabs voice AI to edge devices powered by NVIDIA accelerated computing, potential future applications could include helping workers interact with heavy industrial machinery by voice, supporting hands-free interaction with robots on the factory floor, and enabling operators and drivers to speak naturally with their vehicles. Combining ElevenLabs’ conversational AI technologies with NVIDIA accelerated computing, the collaboration aims to make physical AI more intuitive and accessible.