21.2K ·GitHub
What is this?
It's VCClient, a real-time AI voice changer: it converts your microphone input on the fly so you can speak live with a different voice.
It's VCClient, a real-time AI voice changer: it converts your microphone input on the fly so you can speak live with a different voice.Is this a desktop app, or can it run as a server?Both: it can run entirely on your local PC, or serve the conversion over the network — useful for offloading the conversion load to another machine while a heavy app like a game hogs your PC.
Both: it can run entirely on your local PC, or serve the conversion over the network — useful for offloading the conversion load to another machine while a heavy app like a game hogs your PC.How do I get it running?Download a prebuilt edition for Windows or M1 Mac from the project's Hugging Face repositories; on Linux you clone the repository and run it from there. Google Colab is also supported.
Download a prebuilt edition for Windows or M1 Mac from the project's Hugging Face repositories; on Linux you clone the repository and run it from there. Google Colab is also supported.Any gotchas when running it on macOS?The binaries are not code-signed by the developer, so Gatekeeper shows a warning on first launch; hold Ctrl while clicking the icon to open it, and run it at your own risk.
The binaries are not code-signed by the developer, so Gatekeeper shows a warning on first launch; hold Ctrl while clicking the icon to open it, and run it at your own risk.Is the license a standard open-source one?No standard license is declared (GitHub lists it as NOASSERTION). On top of that, some bundled sample voices carry their own terms — e.g. audio converted with the Tsukuyomi-chan voice must follow that corpus's usage rules.
No standard license is declared (GitHub lists it as NOASSERTION). On top of that, some bundled sample voices carry their own terms — e.g. audio converted with the Tsukuyomi-chan voice must follow that corpus's usage rules.Do I need an NVIDIA GPU to run it?Not always. Which models you can use depends on the edition: std runs Beatrice v2 on Windows, M1 Mac, and Linux (x86-64/aarch64), while RVC needs the cuda edition on Windows or an onnx edition on Windows/M1 Mac.
Not always. Which models you can use depends on the edition: std runs Beatrice v2 on Windows, M1 Mac, and Linux (x86-64/aarch64), while RVC needs the cuda edition on Windows or an onnx edition on Windows/M1 Mac.Which AI voice models does it support?Version 2 runs RVC and Beatrice v2. The older 1.x line additionally supported MMVC, so-vits-svc, DDSP-SVC, and Beatrice v1 (Windows only), and 1.x and 2.x can run side by side on the same machine.
Version 2 runs RVC and Beatrice v2. The older 1.x line additionally supported MMVC, so-vits-svc, DDSP-SVC, and Beatrice v1 (Windows only), and 1.x and 2.x can run side by side on the same machine.What is Beatrice?It's one of the supported voice-conversion model families. The current std edition is built around Beatrice v2, whose license is proprietary, and the author links separate training code for it (a repo plus a Colab version).
It's one of the supported voice-conversion model families. The current std edition is built around Beatrice v2, whose license is proprietary, and the author links separate training code for it (a repo plus a Colab version).Can I control it from my own scripts?Yes — it ships a REST API, so you can operate it with built-in HTTP clients like curl or write a client in any programming language.
Yes — it ships a REST API, so you can operate it with built-in HTTP clients like curl or write a client in any programming language.Is the project still actively developed?Yes — releases come out frequently as betas and alphas. Recent ones added Beatrice v2 support, shortcut keys, buffer visualization, and experimental RTX 5090 support that the author has not been able to verify since he doesn't own that card.
Yes — releases come out frequently as betas and alphas. Recent ones added Beatrice v2 support, shortcut keys, buffer visualization, and experimental RTX 5090 support that the author has not been able to verify since he doesn't own that card.What are similar projects?GPT-SoVITS, MOSS-TTS-Nano, Deep-Live-Cam, voicebox, Retrieval-based-Voice-Conversion-WebUI, so-vits-svc, PDFMathTranslate, video-subtitle-remover
GPT-SoVITS, MOSS-TTS-Nano, Deep-Live-Cam, voicebox, Retrieval-based-Voice-Conversion-WebUI, so-vits-svc, PDFMathTranslate, video-subtitle-removerWhich topics is it in?voice-conversion, voice-ai, japanese, realtime, audio-processing, onnx, training, apple-silicon, audio, cuda, rest-api, gui
voice-conversion, voice-ai, japanese, realtime, audio-processing, onnx, training, apple-silicon, audio, cuda, rest-api, gui