ncnn
Tencent's high-performance neural network inference framework for mobile.
About
ncnn is Tencent's open-source inference framework optimized for mobile, embedded, and desktop deployment. It has no third-party runtime dependencies and runs across CPU and Vulkan GPU backends, with tools like pnnx for converting PyTorch and ONNX models into ncnn format.
It is battle-tested inside Tencent's own products including QQ, Qzone, WeChat, and Pitu. The project is BSD-3 licensed and supports deploying models to phones, PCs, browsers, and edge devices.
What it can do
- Dependency-free mobile inference
- CPU and Vulkan GPU backends
- pnnx model conversion from PyTorch and ONNX
- Deployment to phones, browsers, and edge devices
Setup
Clone the Tencent/ncnn repository and build with CMake, or pull the prebuilt releases for your platform.
Prerequisites
CMake and a C++ toolchain; models converted via pnnx or the ncnn optimizer.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 116; Trending score 74/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.