whisper.cpp

A C/C++ port of OpenAI's Whisper speech recognition model, built for fast offline transcription on many platforms and hardware backends.

Made by ggml.

Licenses

Code: MIT (opens in a new tab)
Read by Robotheca from LICENSE (opens in a new tab) (MIT) at commit 0c7be69 (opens in a new tab), on 16 Oct 2026.
Commercial use: Yes

Runs Whisper automatic speech recognition inference with no external dependencies, implemented in plain C/C++.
Optimized for Apple Silicon with Metal and Core ML, and supports CPU-only inference on x86, ARM and POWER.
Offers GPU and accelerator backends including NVIDIA CUDA, Vulkan, AMD ROCm, OpenVINO, Ascend NPU and AMD Ryzen AI NPU.
Supports integer quantization of models to reduce memory use and disk space.
Provides command-line tools for transcription, real-time microphone streaming, benchmarking and a web server, plus bindings for several languages.
Runs on iOS, Android, Windows, Linux, WebAssembly, Raspberry Pi and in Docker containers.

Written by Robotheca from the project's README (opens in a new tab) at commit 0c7be69.

Difficulty: Intermediate
Hardware it needs, as the project states it: Runs on CPU alone; GPU and accelerator support is optional and needs the matching hardware and driver stack, such as NVIDIA CUDA, Apple Silicon Metal or Core ML, or AMD Ryzen AI NPU. Memory need depends on model size, for example about 273 MB for tiny and about 3.9 GB for large.
How to start: The project's README (opens in a new tab)

Checked on GitHub on 16 Oct 2026.

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.