ElevenLabs releases v4 speech model with enhanced expression control and 90‑language support
The new v4 model can clone a voice from just a ten‑second audio clip and offers finer control over emotional expression.
ElevenLabs has announced the launch of its v4 speech synthesis model, which expands language coverage to ninety languages and introduces more nuanced expression controls.
Unlike earlier versions, the v4 model can create a realistic voice clone using only a ten‑second recording, dramatically lowering the barrier for custom voice creation.
The improved expression controls allow developers to adjust pitch, tone, and emotional nuances in real time, making the output suitable for applications ranging from virtual assistants to audiobooks.
For Uzbekistan’s growing tech sector, the ability to generate localized voice content in Uzbek and other regional languages opens new opportunities in education, media, and accessibility services.
Compared to the previous v3 release, v4 offers higher fidelity and lower latency, addressing two common criticisms of synthetic speech systems.
Industry observers see this update as a step toward more natural‑sounding, multilingual AI voices that can be deployed across a wide range of consumer and enterprise products.
