What happened
ElevenLabs announced a suite of new products and partnerships at its New York Summit, including the Interaction Engine for adaptive AI routing, the ElevenAI enterprise application, and expanded collaborations with NVIDIA and McKinsey.
At its New York Summit, ElevenLabs introduced the Interaction Engine, a system designed to manage the rhythm of AI conversations by handling turn-taking and interruptions. A key of this engine is its ability to dynamically select the appropriate Large Language Model (LLM) for each part of an exchange, using cheaper models for triage and reserving frontier models for complex tasks to optimize token usage and cost.
The company also launched ElevenAI, a new application positioned as an AI teammate for enterprise teams. This tool listens to spoken conversations, including in-person meetings, and converts them into notes and next steps using the Scribe v3 Speech to Text model. It includes built-in notice and consent mechanisms for recording and offers free early access to users.
ElevenLabs announced an expansion of its partnership with NVIDIA, making the Nemotron 3.5 Super VL model available within ElevenAgents. The companies are also adding 15 MW of new Vera Rubin capacity and collaborating on an on-premises voice agent for Poland’s national e-Health platform to assist with appointment scheduling for patients using landlines.
In the enterprise sector, ElevenLabs launched a formal Partner Program and announced an alliance with McKinsey. This collaboration pairs ElevenAgents with McKinsey’s enterprise transformation expertise and its QuantumBlack AI division. The company also introduced the ElevenLabs Marketplace, a curated platform for industry-specific AI applications built by partners, allowing enterprises to browse and connect with these solutions.
Why it matters
These announcements signal a shift from standalone AI models to integrated interaction infrastructure. By introducing the Interaction Engine, ElevenLabs addresses cost and latency challenges in agentic workflows by dynamically selecting between cheaper and frontier models based on task complexity. The launch of ElevenAI extends this capability to internal enterprise operations, offering consent-based meeting transcription and action item generation. Furthermore, the partnership with NVIDIA to integrate Nemotron 3.5 Super VL and add 15 MW of Vera Rubin capacity demonstrates a significant scaling of infrastructure. The alliance with McKinsey and the launch of the Partner Program lower the barrier for enterprise adoption by leveraging existing consulting and integration networks. This move positions ElevenLabs not just as a voice model provider, but as a critical infrastructure layer for human-AI interaction across both customer-facing and internal business processes, potentially reshaping how enterprises deploy conversational AI at scale.
The introduction of the Interaction Engine addresses a critical economic and technical challenge in agentic AI: the high cost and latency associated with using frontier models for every step of a conversation. By automating model selection based on task complexity, ElevenLabs offers a practical path for enterprises to scale conversational AI without prohibitive operational costs.
The launch of ElevenAI marks a strategic expansion from customer-facing voice agents to internal productivity tools. By focusing on consent-based meeting transcription and action item generation, ElevenLabs is targeting the internal workflow of knowledge workers, potentially competing with established meeting assistant tools while leveraging its superior speech-to-text capabilities.
The partnership with NVIDIA is significant for infrastructure scalability. The addition of 15 MW of Vera Rubin capacity and the integration of Nemotron 3.5 Super VL suggest that ElevenLabs is securing the necessary hardware and model resources to support high-volume, low-latency voice interactions, which are more computationally intensive than text-based AI.
The alliance with McKinsey and the launch of the Partner Program indicate a maturing go-to-market strategy. By integrating with major consulting firms and systems integrators, ElevenLabs is reducing the friction for enterprise adoption, allowing companies to deploy complex AI agents through trusted third parties who understand their specific industry contexts and technical stacks.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
What to watch next
Monitor the general availability timeline for the Interaction Engine and the specific performance metrics of the Scribe v3 Medical model. Watch for the rollout of the ElevenLabs Marketplace and how the McKinsey alliance translates into concrete enterprise deployments. Additionally, track the implementation of the on-premises voice agent for Poland’s national e-Health platform as a case study for regulated industry adoption.
The general availability timeline for the Interaction Engine and the specific performance benchmarks of the Scribe v3 Medical model, particularly its accuracy in distinguishing complex medical terminology.
The rollout of the ElevenLabs Marketplace and the types of industry-specific applications that partners will offer, as well as the pricing structure for accessing these agents through existing ElevenLabs commitments.
The implementation details of the on-premises voice agent for Poland’s national e-Health platform, which will serve as a key case study for deploying AI in highly regulated, privacy-sensitive environments.
The extent of the McKinsey alliance, including specific joint solutions or case studies that demonstrate how ElevenAgents can be integrated into broader enterprise transformation initiatives.