VoxCPM
VoxCPM is an AI tool in the Voice & Audio category. Useful for text-to-speech, voice cloning, voice design, open source, local.
Open-source text-to-speech model from OpenBMB that designs a new voice from a written description or clones one from a short clip, in 30 languages.
Modelo de voz aberto da OpenBMB que inventa uma voz a partir de uma descrição escrita ou clona uma a partir de um trecho curto, em 30 línguas.
How it works
A film by Sapiens Sintéticos · 1:15
AITag Review
VoxCPM2 is OpenBMB's open voice model: 2 billion parameters, 30 languages (Portuguese included) and 48 kHz audio, with code and weights under Apache-2.0, free for commercial use. It makes a voice two ways: describe the timbre in a sentence in parentheses, with no recording at all, or hand it a short reference clip and it carries on in that same timbre. It runs on a consumer graphics card (the README puts it around 8 GB), and at Sapiens it is the default voice of the lesson videos we build in code, running locally on our RTX 5060 Ti. Their own rule forbids using the clone to impersonate anyone and asks you to label what is generated.
Where to find it
Features
- Category: Voice & Audio
- Tags:text-to-speechvoice cloningvoice designopen sourcelocalmultilingual
- Website:Visit
- Curated: No



