Likelion AI Research

Korean TTS Engine

Text-to-speech synthesis for the Korean language.
A demonstration by Bebridge for Likelion.

Inference Server

GPU Inference Server

Accelerated L4 GPU inference on AWS EC2

Instance: g6.xlarge

Specs: 4 vCPU, 16GB RAM, NVIDIA L4 GPU

Model Selection

Voice Selection

Reference Audio (Voice Cloning)

Reference Audio File (10-30s)

Reference Text (What is said in the audio?)

Reference Language

Input Text22 chars

Speed1x

Ready to synthesize

Adjust settings and click Generate