# ggml-org/whisper.cpp

Port of OpenAI's Whisper model in C/C++

Repository: https://github.com/ggml-org/whisper.cpp
Canonical: https://ross.abutalabs.com/products/whispercpp
Language: C++
License: MIT
License Family: permissive
Topics: openai, speech-to-text, transformer, whisper, inference, speech-recognition
Last push: 2026-08-25T13:11:32+00:00

## Health v2 (maintenance only)
Score: 99/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 100
- inputs: {"age_days": 1438, "days_push": 8, "days_rel": 13, "gap_med": 18, "n_releases_24m": 20}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 53203, forks 6097 (observed 2026-08-28T04:12:17.895392+00:00)

## What it is
A high-performance C/C++ port of OpenAI's Whisper automatic speech recognition model with no external dependencies. It supports CPU, GPU (Metal, CUDA, Vulkan, ROCm), and NPU inference with quantization, running fully offline across desktop, mobile, and WebAssembly platforms.

## Use cases
- transcribe audio to text offline
- run whisper on-device on iphone or android
- add speech-to-text to a c++ application
- transcribe audio on raspberry pi
- run whisper locally with gpu acceleration
- quantized speech recognition for embedded devices

## When to choose
- you need fast, dependency-free, offline speech-to-text in C/C++ or via bindings
- you want on-device transcription on mobile, desktop, or embedded hardware
- you need GPU or NPU acceleration with quantization for low resource usage

## When to avoid
- you want a simple Python API and don't care about performance or dependencies (use openai/whisper or faster-whisper)
- you need speaker diarization or full audio pipeline features beyond transcription
- you need training or fine-tuning of Whisper models

## Facets
- artifact type: library
- maturity: stable
- function: speech-recognition, llm-inference, machine-learning
- domain: speech-processing, machine-learning
- platform: windows, wasm, cpp, cross-platform
- tags: whisper, asr, speech-to-text, on-device-inference, quantization, ggml, offline-transcription, natural-language-processing, macos, linux, ios, android, gpu

## Member repositories
- ggml-org/whisper.cpp (main) score 99

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:17.895392+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:19:34.440675+00:00, confidence not recorded.
  - readme: https://github.com/ggml-org/whisper.cpp (fetched 2026-08-28T04:12:17.895392+00:00, sha 9678da332d93)
- Data as of 2026-08-30T08:39:29.467469+00:00.