ElevenLabs Introduces 11ai—the First Action-Oriented Voice Assistant

ElevenLabs recently announced the launch of its innovative 11ai voice assistant. Developed by Louis Jordan from the engineering team, the project was unveiled on June 23, 2025, as an alpha version demonstrating the capabilities of the ElevenLabs Conversational AI platform.

11ai differs from traditional voice assistants primarily in that it can actually take action and perform specific tasks within your everyday workflows. While previous voice assistants were largely limited to answering questions and basic conversation, 11ai goes beyond these boundaries through integration with the Model Context Protocol (MCP), allowing it to connect directly to the tools you use every day.

Voice-First Productivity

Traditional voice assistants have long promised to revolutionize how we interact with technology, but their capabilities have remained limited. They could answer questions, but they could not conduct research based on provided data or carry out meaningful actions within workflows. 11ai addresses this issue by connecting directly to commonly used tools through MCP integration.

Thanks to this technology, you can carry out complex tasks using only your voice. For example, in the morning you can say, "Plan my day and add priority tasks to Linear," and the assistant will not only create your schedule but also automatically add the tasks to your project management tool. For customer research, simply say, "Use Perplexity to research today's meeting with a prospective client and summarize their recent funding," and 11ai will perform a complete analysis.

For project management, you can ask, "Search our Linear issues for the API bug and create a new ticket for follow-up work," or for team communication, "Summarize yesterday's messages from the developer Slack channel." 11ai attempts to understand context across your tools and perform sequential actions. When you request customer research, the goal is to search connected systems, find relevant data, and potentially act on that information by updating the CRM system or sending a team message.

Model Context Protocol Integration

The Model Context Protocol (MCP) provides a standardized way for AI assistants to integrate with external APIs using a unified protocol. ElevenLabs Conversational AI now includes built-in support for MCP, enabling AI agents to connect seamlessly to services such as Salesforce, HubSpot, Gmail, Zapier, and many others.

The platform offers prebuilt integrations for key tools. Perplexity provides research capabilities using real-time web data, Linear handles issue tracking and project management, Slack enables team communication and message management, while Notion takes care of task and knowledge management. Additional integrations are released every week, continuously expanding the platform's capabilities.

In addition to these prebuilt integrations, 11ai also supports custom MCP servers. If your team has built internal tools or uses specialized software, you can connect custom MCP servers and extend 11ai's capabilities into your unique workflow. The MCP architecture ensures that all connections are secure and that 11ai can access only the specific actions you have authorized. Each integration can be configured with the appropriate permissions, giving you full control over what your AI assistant can and cannot do.

ElevenLabs Conversational AI Technology

11ai serves as a proof of concept demonstrating what developers can build using ElevenLabs Conversational AI—a low-latency platform for creating scalable voice agents. This demonstration shows how the underlying technology can be used to create sophisticated voice applications with real-world integrations.

The platform's key features include ultra-low latency for real-time conversations with minimal delay, multimodal support for both voice and text interactions within the same session, integrated RAG (Retrieval-Augmented Generation) for accessing external knowledge bases with context-aware responses, automatic language detection for seamless multilingual conversations, and enterprise readiness with HIPAA compliance and enterprise-grade security.

11ai shows how these capabilities combine to create natural, productive voice interactions. The platform handles the complex orchestration of speech processing, intent understanding, tool integration, and response generation—all while maintaining a conversational flow that makes voice interaction feel natural.

Users can choose from more than 5,000 voices in the library or create a custom voice clone that sounds like them. This personalization makes your AI assistant feel like a natural extension of your workflow rather than a generic bot.

Availability and Pricing

11ai has been available as an alpha proof of concept since its launch. During this experimental phase, the company is offering free access to gather feedback and demonstrate the potential of voice-first AI assistants.

According to available information about ElevenLabs pricing, the platform generally offers several service tiers. The free plan provides 10,000 credits per month (approximately 10 minutes of audio), making it suitable for basic testing. Paid plans start at $5 per month with 30,000 credits, followed by the Creator plan at $22 with 100,000 credits, and go up to an enterprise solution at $1,320 with 11 million credits per month.

To get started with 11ai, you need to register at 11.ai, complete the setup process, and choose a voice from the library of more than 5,000 voices or create a custom voice clone. You can then connect your tools, including Google Calendar, Linear, Slack, Perplexity, and any custom MCP servers, and begin your first conversation by trying basic workflows such as planning your day or researching a topic.

During the alpha period, the company is particularly interested in feedback on which integrations are most valuable to your workflow, which additional MCP servers you would like to see supported natively, how voice interaction feels compared with traditional interfaces, and what new capabilities would make 11ai essential to your daily routine.

The Czech language is fully supported across the ElevenLabs platform, including advanced features such as voice cloning and adaptation to regional accents. The service is accessible online, and anyone can register and start using Czech TTS features directly from the ElevenLabs platform. There are no indications of regional restrictions for Czech users—all features, including pricing and credits, appear to be universally available.

As the company gathers feedback and iterates on the platform, it plans to add new integrations, improve conversational flow, and expand the actions that 11ai can perform on your behalf. As a proof of concept, 11ai demonstrates a vision for the future of human-computer interaction: natural conversation that leads to meaningful action.