ElevenLabs has announced that developers can now build conversational agents powered by generative Artificial Intelligence (AI) on the platform, which have customizable features and are compatible with Gemini, GPT and Claude.
This software startup uses generative AI focused on voice-related issues, such as cloning and text-to-speech, and aims to eliminate linguistic barriers to content.
The firm, which already has an AI dubbing tool and a reading application with voices of classic film actors, among other features, has announced that it has made conversational AI agents available to users.
This is a feature that some users had already been able to test, but that can now be used by all people interested in building these ‘bots’, customizing both your tone of voice and the length of your responsesamong other variables.
In developing these agents, ElevenLabs has encountered greater difficulty in integrating the knowledge base and managing customer outages, the company’s head of growth, Sam Sklar, confirmed to TechCrunch.
For this reason, the firm has decided to create a specific channel so that developers can build these ‘bots’, which makes their configuration and use easier. Once you have logged in to the user account, you can choose a main language and a specific message to personalize the ‘chatbot’ experience.
Developers also have to select a large language model (LLM), this is Google’s Gemini, OpenaAI’s GPT, or Anthropic’s Claude; as well as the level of creativity of the answers and the token usage limit.
Other configurable options are voice, latency, stability, authentication criteria and the maximum duration of the conversation with the artificial intelligence agent.
On the other hand, users have the possibility to add their own knowledge base to power the agent, such as a url, a block of text or a file; as well as your own personalized LLM.
In this sense, it is worth remembering that the ElevenLabs software development kit (SDK) is compatible con Python, JavaScript, React y Swift. Additionally, for further customization, the company offers the WebSocket application programming interface (API).