ElevenLabs, the innovative startup renowned for its AI voice cloning and text-to-speech solutions, has officially launched a groundbreaking feature that allows users to create sophisticated conversational AI agents. This exciting development was revealed on Monday, marking a significant milestone in the company's evolution.

The new capability enables users to build fully customizable conversational agents tailored to their specific needs. Users can modify essential attributes such as tone of voice, response length, and even the agent’s unique persona by selecting the primary language, initial greetings, and system prompts. This feature is set to empower businesses and developers to create seamless customer interactions that feel organic and engaging.

According to Sam Sklar, ElevenLabs’ head of growth, the demand for conversational AI agents among their clients has spurred this enhancement. Many clients were already integrating voice capabilities for AI bots, but faced challenges like knowledge base integration and customer interaction handling. To address these issues, ElevenLabs has developed a comprehensive pipeline dedicated to building robust conversational bots.

Developers using ElevenLabs’ platform have the option to choose from popular large language models (LLMs) such as Gemini, GPT, or Claude. They can also adjust response creativity through a “temperature” setting and restrict token usage to manage costs effectively. Additionally, features such as voice customization, latency adjustments, stability settings, and maximum conversation lengths grant developers a high level of control over their AI agents.

What’s more, users can enrich their bots by incorporating their own knowledge bases — whether it’s a file, URL, or a block of text — and connect their custom LLMs directly to the AI. The ElevenLabs SDK supports multiple programming languages, including Python, Javascript, React, and Swift, offering flexibility for developers. A WebSocket API further enhances customization options for tech-savvy companies.

Moreover, ElevenLabs allows companies to set parameters for data collection — capturing vital information like customer names and emails during interactions, and establishing evaluation criteria in natural language to gauge the performance of their conversational agents.

While ElevenLabs currently leverages its existing framework for text-to-speech, the company is actively developing speech-to-text capabilities, a crucial component of their future offerings. Though not currently available as a standalone product, this move could soon position ElevenLabs as a competitor in the speech-to-text market, going head-to-head with industry giants like Google, Microsoft, and Amazon, as well as specialized services like OpenAI's Whisper, AssemblyAI, and others.

In an age where conversational AI is becoming fundamentally integral to customer service and engagement strategies, ElevenLabs’ new platform is poised to set a new standard in the industry, enabling businesses to create more personalized and efficient interactions than ever before.