WebGPU.Studio

Text to Speech

Convert text to natural-sounding speech using WebGPU

Model Setup

Idle
LFM2.5 Audio is a large model (~1.5 GB). Click load to start a one-time download to browser cache.

Liquid multimodal audio model with high-quality TTS

LFM2.5 Audio (Q4) voice profile: UK female voice. This model also supports ASR and interleaved voice turns in the new LFM Audio demo.
Powered byTransformers.js&LFM2.5 Audio (Q4)

All processing happens locally in your browser