# Tencent/ncnn

ncnn is a high-performance neural network inference framework optimized for the mobile platform

Repository: https://github.com/Tencent/ncnn
Canonical: https://ross.abutalabs.com/products/ncnn
Language: C++
License: NOASSERTION
License Family: other
Topics: inference, high-preformance, simd, arm-neon, deep-learning, artificial-intelligence, android, ios, ncnn, vulkan, neural-network, caffe, mxnet, pytorch, onnx, darknet, tensorflow, mlir, keras, riscv
Last push: 2026-08-25T03:21:15+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 62, longevity 100
- inputs: {"age_days": 3351, "days_push": 8, "days_rel": 99, "gap_med": 123, "n_releases_24m": 6}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 23753, forks 4492 (observed 2026-08-28T04:11:35.325004+00:00)

## What it is
ncnn is a high-performance neural network inference framework written in C++ and optimized for mobile, embedded, and desktop deployment. It has no third-party runtime dependencies, supports CPU and Vulkan GPU backends, and includes tools like pnnx for converting PyTorch and ONNX models.

## Use cases
- run deep learning models on android and ios apps
- deploy neural network inference on edge devices without dependencies
- convert pytorch or onnx models to a mobile-friendly format
- accelerate inference with vulkan gpu on phones and pcs
- run computer vision models like object detection on mobile
- embed neural network inference in c++ applications

## When to choose
- you need fast on-device inference on mobile or embedded hardware
- you want a dependency-free c++ inference runtime
- you need cpu (arm neon) and vulkan gpu support
- you are deploying models in production mobile apps

## When to avoid
- you need training or fine-tuning, not just inference
- you want a python-first workflow with rich ecosystem tooling
- you need large transformer/llm serving features out of the box
- you prefer a framework with extensive prebuilt model zoo and cloud integration

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, llm-inference, computer-vision, image-processing, gpu-computing
- domain: deep-learning, artificial-intelligence, mobile-development, embedded-systems, cross-platform
- platform: cpp, cross-platform, wasm, windows, embedded
- tags: neural-network-inference, vulkan, arm-neon, simd, model-conversion, pnnx, onnx, pytorch, edge-deployment, mobile-ai, android, ios, mobile, desktop, linux, macos, gpu

## Member repositories
- Tencent/ncnn (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:35.325004+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 2026-08-29T16:56:21.245077+00:00, confidence not recorded.
  - readme: https://github.com/Tencent/ncnn (fetched 2026-08-28T04:11:35.325004+00:00, sha 3038d2dd648e)
- Data as of 2026-08-30T08:39:29.467469+00:00.