ElevenLabs v3

ElevenLabs[audio]

Latest model — natural multi-speaker dialogue with audio tags for emotions, reactions, and pacing. 70+ languages.

Max Output

5K tokens

Released

2025-06

Features

[multi-speaker][audio-tags][emotion-control][70+-languages][natural-dialogue]

Technical Specifications

Languages70

The Lab

location: New York, New York, USAfounded: 2022backed by: Andreessen Horowitz, Sequoia Capital; venture-backed unicorn

Founded in 2022 and headquartered in New York, ElevenLabs is an AI voice and audio research company backed by Andreessen Horowitz and Sequoia Capital. The lab develops speech synthesis, voice cloning, and conversational AI-agent platforms used broadly in content creation, accessibility, and gaming. Its voice models set the quality benchmark for synthetic speech, and its voice-clone consent policy is a de facto industry standard for the audio-rights layer.

more about ElevenLabs →

Model ID: eleven_v3