MobileNetV4 - Universal Models for the Mobile Ecosystem

Danfeng Qin, Chas Leichner, M. Delakis, M. Fornoni, Shixin Luo +9 more
2/14/2026

No official code found. Best implementation: tensorflow/models (77,659 stars, last commit 2026-04-29). 1 more below.

Abstract

We present the latest generation of MobileNets, known as MobileNetV4 (MNv4), featuring universally efficient architecture designs for mobile devices. At its core, we introduce the Universal Inverted Bottleneck (UIB) search block, a unified and flexible structure that merges Inverted Bottleneck (IB), ConvNext, Feed Forward Network (FFN), and a novel Extra Depthwise (ExtraDW) variant. Alongside UIB, we present Mobile MQA, an attention block tailored for mobile accelerators, delivering a significant 39% speedup. An optimized neural architecture search (NAS) recipe is also introduced which improves MNv4 search effectiveness. The integration of UIB, Mobile MQA and the refined NAS recipe results in a new suite of MNv4 models that are mostly Pareto optimal across mobile CPUs, DSPs, GPUs, as well as specialized accelerators like Apple Neural Engine and Google Pixel EdgeTPU - a characteristic not found in any other models tested. Finally, to further boost accuracy, we introduce a novel distillation technique. Enhanced by this technique, our MNv4-Hybrid-Large model delivers 87% ImageNet-1K accuracy, with a Pixel 8 EdgeTPU runtime of just 3.8ms.

View on arXivView PDF

Code Implementations(2)

tensorflow/modelsCommunity100%

Models and examples built with TensorFlow

77,65944,853C++, CSSFeb 5, 20165 months agoNOASSERTION
jpanagos/MobileNetV4-PyTorchCommunity57%
GPL-3.0

Straightforward implementation of MobileNetV4 https://arxiv.org/abs/2404.10518

10PythonMay 31, 20241 years agoGPL-3.0

Similar Papers

DNA Family: Boosting Weight-Sharing NAS with Block-Wise Supervisions

81% match1 repo2342024

FastBoost: Progressive Attention with Dynamic Scaling for Efficient Deep Learning

81% match2025

Puzzle: Distillation-Based NAS for Inference-Optimized LLMs

81% match1 repo42026

Neuralink: Fast LLM Inference on Smartphones with Neuron Co-Activation Linking

81% match2024

Differentiable architecture search with multi-dimensional attention for spiking neural networks

80% match2024

SHViT: Single-Head Vision Transformer with Memory Efficient Macro Design

80% match2 repos1442026

Cite this paper

@article{qin2026mobilenetv,
  title  = {MobileNetV4 - Universal Models for the Mobile Ecosystem},
  author = {Danfeng Qin and Chas Leichner and M. Delakis and M. Fornoni and Shixin Luo and Fan Yang and Weijun Wang and Colby R. Banbury and Chengxi Ye and Berkin Akin and Vaibhav Aggarwal and Tenghui Zhu and Daniele Moro and Andrew Howard},
  year   = {2026},
  eprint = {2404.10518},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/2404.10518},
  journal = {ECCV 2024 2024}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2404.10518.svg)](https://codeofpaper.com/paper/2404.10518)

Discussion