GPT-SoVITS
An open-source AI voice synthesis tool that supports few-shot voice cloning by combining GPT and SoVITS technologies to generate speech from minimal audio samples.
Timeline (last 90 days)
IboChat's /speak endpoint tries three voice paths in order: GPT-SoVITS (best voice cloning, takes about 5–18 seconds, supports only English, Chinese, Japanese, and Korean, and permanently occupies about 3.9GB of VRAM once loaded), OmniVoice Studio (supports any language, takes about 30–40 seconds, called via an OpenAI-compatible endpoint), and Edge TTS (Microsoft's free neural voices, no cloning capability, supports 24 languages, zero latency, never fails).
GPT-SoVITS is an open-source few-shot voice cloning and synthesis project that combines GPT-style autoregressive modeling with the SoVITS acoustic model; it can clone a timbre from only a few seconds to a few tens of seconds of target audio, and its official pretrained models mainly cover languages such as Chinese, English, Japanese, and Korean.
IboChat runs on a single GPU with 4GB of VRAM, with GPT-SoVITS and OmniVoice sharing the same GPU; the author solves the audio corruption caused by resource contention by starting GPT-SoVITS on demand, using a watchdog thread to kill the process after 120 seconds of idleness, and clearing OmniVoice's VRAM before each call.
All Facts (5)
ASR市场支持创作者上架AI整合包，如Stable Diffusion WebUI、ComfyUI、GPT-SoVITS等
85%UnverifiedModern neural network-based TTS models (such as Tacotron, VITS, GPT-SoVITS, etc.) can generate natural speech close to that of a real person, and support emotion control and voice timbre cloning.
60%UnverifiedIboChat's /speak endpoint tries three voice paths in order: GPT-SoVITS (best voice cloning, takes about 5–18 seconds, supports only English, Chinese, Japanese, and Korean, and permanently occupies about 3.9GB of VRAM once loaded), OmniVoice Studio (supports any language, takes about 30–40 seconds, called via an OpenAI-compatible endpoint), and Edge TTS (Microsoft's free neural voices, no cloning capability, supports 24 languages, zero latency, never fails).
50%UnverifiedGPT-SoVITS is an open-source few-shot voice cloning and synthesis project that combines GPT-style autoregressive modeling with the SoVITS acoustic model; it can clone a timbre from only a few seconds to a few tens of seconds of target audio, and its official pretrained models mainly cover languages such as Chinese, English, Japanese, and Korean.
50%UnverifiedIboChat runs on a single GPU with 4GB of VRAM, with GPT-SoVITS and OmniVoice sharing the same GPU; the author solves the audio corruption caused by resource contention by starting GPT-SoVITS on demand, using a watchdog thread to kill the process after 120 seconds of idleness, and clearing OmniVoice's VRAM before each call.
50%