Generate Speech from Text with ElevenLabs

The Elevenlabs.io Tool enables LlamaIndex agents to convert text into speech audio files using the Elevenlabs.io text-to-speech API. It provides two core functions: synthesizing audio from input strings and retrieving available voice models.

Use this tool when you want to integrate Elevenlabs.io text-to-speech capabilities into LlamaIndex agents.

Consult the Elevenlabs.io documentation and your specific requirements to determine if this integration approach fits your needs.

Users provide an Elevenlabs.io API key during initialization and text strings for conversion. The text_to_speech function takes an input string and synthesizes audio to play on the users computer. The get_voices function returns Pydantic models listing all available Elevenlabs.io voices.

This tool integrates with LlamaIndex's FunctionAgent and agent workflow system, allowing it to be loaded as part of an agent's tool list. It works alongside LlamaIndex LLMs such as OpenAI models (e.g., gpt-4.1). The tool is designed to be used as a way to load data as a Tool in a Agent.

from llama_index.tools.elevenlabs import ElevenLabsToolSpec
from llama_index.core.agent.workflow import FunctionAgent
from llama_index.llms.openai import OpenAI

speech_tool = ElevenLabsToolSpec(api_key="your-key")

agent = FunctionAgent(
    tools=speech_tool.to_tool_list(),
    llm=OpenAI(model="gpt-4.1"),
)
print(
    await agent.run(
        'Create speech from the following text "Hello world!" and output the file to "speech.wav"'
    )
)

This tool is for developers working with LlamaIndex agents who want to add Elevenlabs.io text-to-speech functionality. It abstracts Elevenlabs.io API complexity into agent-compatible functions for developers familiar with LlamaIndex's agent framework.