Skip to content
UI Automation Framework for Games and Apps
Branch: master
Clone or download
Latest commit ec4d93a Nov 16, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
.github/ISSUE_TEMPLATE add issue template Sep 19, 2018
airtest 完善windows模块和android模块的中文翻译,修正了部分代码注释中的拼写错误 Oct 9, 2019
benchmark fix: English version correction. Apr 15, 2019
docs Update Nov 16, 2019
playground fix script & log dir settings Feb 13, 2019
.coveragerc add aircv/win in coverage; fix test_win Feb 8, 2018
.gitattributes add android/static into linguist-vendored Mar 26, 2018
.travis.yml exclude *.air flake8 check in travis.yml Mar 30, 2018
LICENSE opensource init version Jan 24, 2018 修复在低版本pip下可能出现的路径以/结尾的安装失败问题 Mar 15, 2019 Update Oct 12, 2019 update readme Feb 14, 2019
demo.gif add gif Jan 30, 2018
requirements.txt iOS-Tagent暂不支持最新版facebook-wda Sep 27, 2019 level up to v1.1.1 Sep 29, 2019

Airtest · Build status

Cross-Platform UI Automation Framework for Games and Apps




  • Write Once, Run Anywhere: Airtest provides cross-platform APIs, including app installation, simulated input, assertion and so forth. Airtest uses image recognition technology to locate UI elements so that you can automate games and apps without injecting any code.

  • Fully Scalable: Airtest cases can be easily run on large device farms, using commandline or python API. HTML reports with detailed info and screen recording allow you to quickly locate failure points. NetEase builds Airlab on top of the Airtest Project.

  • AirtestIDE: AirtestIDE is an out of the box GUI tool that helps to create and run cases in a user-friendly way. AirtestIDE supports a complete automation workflow: create -> run -> report.

  • Poco: Poco adds the ability to directly access object(UI widget) hierarchy across the major platforms and game engines. It allows writing instructions in Python, to achieve more advanced automation.

Get started from airtest homepage

Supported Platforms

Android Emulator iOS Windows Unity Cocos2dx Egret WeChat


Use pip to install the Airtest python library.

pip install -U airtest

On MacOS/Linux platform, you need to grant adb execute permission.

# for mac
cd {your_python_path}/site-packages/airtest/core/android/static/adb/mac
# for linux
# cd {your_python_path}/site-packages/airtest/core/android/static/adb/linux
chmod +x adb

Download AirtestIDE from our homepage if you need to use the GUI tool.


You can find the complete Airtest documentation on readthedocs.


Airtest aims at providing platform-independent API so that you can write automated cases once and run it on multiple devices and platforms.

  1. Using connect_device API you can connect to any android/iOS device or windows application.
  2. Then perform simulated input to automate your game or app.
  3. DO NOT forget to make assertions of the expected result.
from airtest.core.api import *

# connect an android phone with adb
# or use connect_device api
# connect_device("Android:///")

swipe(Template("slide_start.png"), Template("slide_end.png"))

For more detailed info, please refer to Airtest Python API reference or take a look at API code

Running .air cases from CLI

Using AirtestIDE, you can easily create automated cases as .air directories. Airtest CLI provides the possibility to execute cases on different host machines and target device platforms without using AirtestIDE itself.

# run cases targeting on Android phone connected to your host machine via ADB
airtest run "path to your .air dir" --device Android:///

# run cases targeting on Windows application whose title matches Unity.*
airtest run "path to your .air dir" --device "Windows:///?title_re=Unity.*"

# generate HTML report after running cases
airtest report "path to your .air dir"

# or use as a python module
python -m airtest run "path to your .air dir" --device Android:///

Try running provided example case: airtest/playground/test_blackjack.air and see Usage of CLI. Here is a multi-device runner sample.


Pull requests are very welcome. Help needed here


Thanks for all these great works that make this project better.

You can’t perform that action at this time.