ElevenLabs and NVIDIA widen their partnership: Nemotron joins ElevenAgents as voice agents head for national-scale deployment

ElevenAgents, ElevenLabs' conversational agent platform, can now draw on NVIDIA's Nemotron open models, letting enterprises combine a reasoning model with ElevenLabs' low-latency multilingual voice models as a system of models for building voice agents. The expanded collaboration brings NVIDIA's models, infrastructure and secure deployment into the ElevenAgents platform, and Nemotron was not previously among the models available there. On the ElevenAgents benchmark for realistic customer-support tasks, Nemotron 3.5 Super VL achieved leading end-to-end task-completion accuracy, with strong performance in financial calculations and billing analysis, at 70% lower cost than other leading open models.

ElevenLabs plans to add 15 MW of next-generation NVIDIA Vera Rubin capacity on top of its existing NVIDIA Blackwell Ultra systems, supporting large-scale deployments with strict security and data-residency requirements and giving its research teams access to optimized compute. The added capacity responds to growing demand from the company's largest customers rather than replacing the existing infrastructure. The companies have worked together since 2022: ElevenLabs' first text-to-speech model was trained on NVIDIA GPUs, NVIDIA has used ElevenLabs technology since 2024 to generate voiceovers for 50+ videos across all major NVIDIA events, and in 2025 NVIDIA became a strategic investor in ElevenLabs.

In Poland, ElevenLabs models and NVIDIA infrastructure power the voice agent for Centralna e-Rejestracja, the central appointment system within the national e-Health platform, with EasyCall supporting delivery; the agent calls patients who only have landlines to remind them of appointments and help them reschedule. The deployment runs on-premises with NVIDIA Confidential Computing, keeping citizen data and audio within government infrastructure and protected during processing, while local deployment reduces latency for real-time public services. Poland's national e-Health platform handles 40 million appointments a year with no-show rates of 10 to 20 percent, so even a small reduction in missed visits can significantly improve access to care.

The demonstrated scope covers model access on the ElevenAgents platform, the comparison on the ElevenAgents customer-support benchmark, the planned 15 MW of Vera Rubin capacity, and the on-premises deployment in Poland's central appointment system; on-device voice AI for robots, vehicles and industrial machinery is described as a potential future application.

The benchmark result applies only to the evaluation of realistic customer-support tasks on the ElevenAgents platform, and the 70% lower cost is a comparison against other leading open models; the Polish deployment's benefit is framed by 40 million appointments a year and 10 to 20 percent no-show rates, with no actual reduction reported; on-device voice AI applications are described as possible directions.