@Honcia13: Open-source TTS is going crazy! New weapons for industrial park scams? Tsinghua OpenBMB just released VoxCPM2: 20 billion parameters + 2 million hours of multilingual data training, 48kHz studio-quality sound! The most intense part is—no Tokenizer needed at all, performing diffusion autoregression directly in continuous latent space, maximizing detail retention!
Summary
Tsinghua University's OpenBMB has released VoxCPM2, an open-source multilingual TTS model with 20 billion parameters. It supports continuous latent space diffusion autoregressive generation without a Tokenizer, offering 48kHz studio-quality audio and powerful voice cloning and design capabilities.
View Cached Full Text
Cached at: 05/12/26, 07:35 AM
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
English | Chinese
👋 Join our community for discussion and support! Feishu | Discord
ModelBest THUHCSI
Similar Articles
@FakeMaidenMaker: Explosive! This open-source project converts text to human-like voice for free, can clone anyone's voice, and adjust timbre with text! GitHub has garnered 30K stars, from Mianbao Intelligent OpenBMB, VoxCPM previously topped both GitHub and HuggingFace charts. Do...
VoxCPM2 is an open-source speech synthesis model from OpenBMB, using a tokenizer-free diffusion autoregressive architecture, supporting 30 languages, voice design, and controllable voice cloning. It can clone a voice with just one sentence, or create a brand new voice using text, outputting 48kHz high-quality audio, and is commercially usable.
@QT9277: "No way, AI voice synthesis has gotten this insane???" I was browsing GitHub today and was completely stunned. VoxCPM2, trending #1, over 20k stars, blowing up overseas. I thought it was another PPT open-source project, but after carefully checking the demo—my ears really couldn't tell which one was real. …
Introducing VoxCPM2, a completely free for commercial use, open-source multilingual voice synthesis model supporting voice design, cloning, and 48kHz high-quality output, ranked #1 on GitHub trending.
@NFTCPS: Fraud call centers have a new weapon — voice cloning has been pushed to new heights again. LuxTTS, a lightweight TTS model, after seeing it I can only say: truly insane. Fast: 150x real-time on a single GPU, even runs faster than real speech on CPU. Clear: 48kHz directly, most models are still stuck at 24kHz…
LuxTTS is a lightweight voice cloning TTS model, supporting 48kHz high-fidelity output, achieving 150x real-time speed on a single GPU, requiring only 1GB VRAM for local operation, with performance comparable to models ten times its size.
OpenBMB/VoxCPM
OpenBMB releases VoxCPM2, a 2B-parameter tokenizer-free TTS model trained on 2M+ hours of multilingual speech data, supporting 30 languages, voice design, controllable cloning, and 48kHz output.
@hisevenih: The AI voice community is blown away. This GitHub open-source black tech takes AI voice to an insane level, truly achieving: one sentence, one voice. Remember this project name: VoxCPM2. It has already gained 20K stars on GitHub. Most incredibly, it doesn't even need a reference audio…
GitHub open-source project VoxCPM2 achieves AI voice cloning without reference audio, generating target voice precisely with just one sentence, has gained 20K stars.