# XiaoMi/mace

MACE is a deep learning inference framework optimized for mobile heterogeneous computing platforms.

Repository: https://github.com/XiaoMi/mace
Canonical: https://ross.abutalabs.com/products/mace
Language: C++
License: Apache-2.0
License Family: permissive
Topics: deep-learning, neural-network, opencl, neon, hvx, machine-learning
Last push: 2024-06-17T09:17:33+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2989, "days_push": 807, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5046, forks 819 (observed 2026-08-28T04:09:08.891630+00:00)

## What it is
MACE (Mobile AI Compute Engine) is a deep learning inference framework optimized for mobile heterogeneous computing on Android, iOS, Linux and Windows devices. It supports TensorFlow, Caffe and ONNX models with runtimes optimized for CPU (NEON), GPU (OpenCL) and DSP (Hexagon).

## Use cases
- run neural network inference on android phones
- deploy tensorflow models on mobile devices
- convert onnx models for mobile inference
- accelerate cnn inference with mobile gpu via opencl
- run models on hexagon dsp
- protect model weights in a mobile app
- reduce memory footprint of on-device inference

## When to choose
- you need fast on-device inference on ARM mobile chips
- you want GPU or DSP acceleration for neural networks on Android
- model protection and small binary size matter
- you need to run TensorFlow, Caffe or ONNX models on mobile

## When to avoid
- you need training, not inference
- you target only server or desktop GPUs
- you need broad support for the latest transformer architectures
- you want a framework with frequent recent updates

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-inference
- domain: deep-learning, mobile-development, machine-learning
- platform: windows, cross-platform, cpp
- tags: mobile-inference, opencl, neon, hexagon, onnx, tensorflow, caffe, model-protection, android, ios, linux

## Member repositories
- XiaoMi/mace (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:08.891630+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 2026-08-29T18:17:22.726149+00:00, confidence not recorded.
  - readme: https://github.com/XiaoMi/mace (fetched 2026-08-28T04:09:08.891630+00:00, sha 857db27dd0d4)
- Data as of 2026-08-30T08:39:29.467469+00:00.