请支持 audio.cpp 作为本地 ASR + TTS/语音克隆后端

Hello everyone! I am the maintainer of audio.cpp. I noticed that the dubbing process in VideoLingo requires both ASR and TTS, so I was wondering if audio.cpp could serve as a unified local backend that covers both parts. Currently, audio.cpp supports more than 70 models, including 15 ASR models and 38 TTS/sound clone models, and is continuously expanding through community contributions. Some of the ASR integrations are contributed directly by the official model teams, including Mistral AI, FunASR, IBM, and Microsoft teams. For TTS, we currently support models such as Qwen3-TTS, IndexTTS 2/2.5, Higgs Audio TTS v3, Fish Audio, PocketTTS, and VibeVoice Breeze TTS2. These models can be run locally through the same C++ runtime and are optimized for CUDA, Metal, Vulkan, and HIP. The inference in audio.cpp is completely independent of Python and provides both Server API and CLI, making it possible for VideoLingo to access a variety of local ASR, TTS, and sound clone models through a unified backend. I wonder if you would be interested in this integration approach. I'm happy to answer any questions you have about audio.cpp.