whisper.cpp

Why we included this project

whisper.cpp is a C/C++ port of OpenAI's Whisper that runs the speech-to-text model locally, with no Python runtime and no heavy dependencies. That makes it a natural fit for embedding transcription into mobile apps, desktop tools, and edge devices that need a small, self-contained binary. It runs on a Raspberry Pi or an iPhone as readily as on an x86 server, with CPU-only inference, integer quantization, and GPU backends for Metal, CUDA, Vulkan, and ROCm. The C-style API, built-in voice activity detection, and ready-made examples also give you a solid base for building an offline voice assistant.