VoxCPM 1.5, developed by FaceWall Intelligence, is open source, further upgrading its voice generation capabilities.

VoxCPM version 1.5 has been officially released as open source. The AudioVAE sampling rate has been increased from 16kHz to 44.1kHz, significantly improving the detail richness of timbre cloning. Generation efficiency has been doubled, requiring only 6.25 tokens to generate 1 second of audio. Additionally, LoRA and full fine-tuning scripts have been added to support deep customization, model stability has been optimized, and audio artifacts have been reduced.