#2929th of 37 in ML frameworks and libraries

ncnn

C++ neural network inference library for mobile, embedded and desktop devices

Stars: 23.9k
License: custom license
Last commit: Oct 2026
Last release: May 2026
Language: C++

Missed the latest GitHub check; these facts may be older.

Overview

ncnn runs neural network inference on CPU or Vulkan GPU with no third-party runtime dependencies. It ships prebuilt packages for Android, iOS, HarmonyOS, macOS, Linux, Windows and WebAssembly, plus cross-compile guides for ARM, MIPS, RISC-V and LoongArch boards. The pnnx tool converts PyTorch and ONNX models to ncnn's param/bin format, and C++ and Python APIs are provided.

Who it is for: Developers deploying trained models to phones, embedded boards and browsers

Strengths

No third-party runtime dependencies; CPU and Vulkan GPU backends
Prebuilt packages for Android, iOS, HarmonyOS, Windows, Linux, macOS and WebAssembly
Builds for many architectures: ARM, MIPS, ppc64, riscv64, loongarch64
pnnx converts PyTorch and ONNX models; C++ and Python APIs

Weaknesses

Inference only; training is not covered
Models must be converted to ncnn param/bin format first
No HTTP serving layer; it is an embedded library
License detected as NOASSERTION; README badge says BSD 3-Clause

What it needs

GPU optional
Models: PyTorch (via pnnx), ONNX (via pnnx), ncnn param/bin

Also in ML frameworks and libraries

See all 37

Also in ML frameworks and libraries
Rank	Project	Score
1	transformersPython library defining open models for text, vision, audio and multimodal tasks167.7k starsApache-2.0	95 out of 100
2	pytorchPython tensor library with GPU support and tape-based autograd for deep learning104.7k starscustom license	94 out of 100
3	ultralyticsPython library and CLI for training and running YOLO vision models62.5k starsAGPL-3.0	90 out of 100
#4	supervisionPython toolkit for loading, annotating and counting computer vision detections Live demo ↗ (opens in a new tab)51.2k starsMIT	87 out of 100
#5	pytorch-image-modelsPyTorch library of image models, pretrained weights and reference training scripts37.2k starsApache-2.0	86 out of 100

How to choose

Pick the framework your target models and papers already use.
Check hardware support (CUDA, ROCm, Apple Silicon, TPU) for your machines.
Look at release cadence; a framework that stalls strands your code.

Health

Releases: Not checked
Top contributor: Not checked
Security advisories: Not checked
Package advisories: Not checked
OpenSSF Scorecard: Not checked

Not read for this project yet: each refresh reads the facts of a few projects. Bots and coding-agent accounts are left out of the commit counts. Not part of the score: how we check.

License in plain words

custom license Custom terms

Use it at work: Check
Change it and rebrand it: Check
Sell it or host it for others: Check

Your part: GitHub could not match these terms to a standard license. Read the license before you use it at work, rebrand it or host it for others.

A plain summary, not legal advice. Read the license on GitHub ↗ (opens in a new tab)

Where to run it

Vast.ai ↗ (opens in a new tab)Rent a GPU by the hour from a marketplace of hosts.

Referral links: if you sign up through one, we get a commission. They never change a rank.

#29 of 37 in ML frameworks and libraries on Archestack, as of 15 Oct 2026.

In your README

The badge updates twice a day.

[![Archestack](https://archestack.com/badge/tencent/ncnn.svg)](https://archestack.com/best-of-ai/projects/ncnn/)

How badges work

On your site or blog

A card with the rank, score and signals, in light or dark to match the reader.

<iframe src="https://archestack.com/embed/best-of-ai/ncnn/" title="ncnn on Archestack" width="400" height="270" style="border:0;max-width:100%" loading="lazy"></iframe>

Preview the card

Best of Open-Source AI on GitHub

Each review comes from the project's README and facts we check on GitHub. The fix form opens with ncnn filled in.