# Software Experience By [wangmingqing](https://paragraph.com/@wangmingqing) ยท 2024-07-31 software experience, software, computer --- There are many AI projects on the market now, and many people use open source projects to build their own projects, and then package them into clients for payment. Is there any software that aggregates these projects together, and is it an open-source and free project? After searching, there is indeed such a tool. Let's take a look at the AI productivity tool codenamed APT that we experienced today. Its introduction is: free and open-source, improving user productivity, and ensuring privacy and data security. Provide efficient and convenient AI solutions, including but not limited to: built-in exclusive ChatGPT, one click batch intelligent processing of images and videos, etc. It looks pretty good. The features mentioned in the project are: The current version number of the project is V2.2.2.0, which has two versions, one is Lite 1.23GB and the other is Full 3.2GB. The Full version adds ChatGPT functionality. You can download according to the desired function, unzip it, and then open the startup file to use it. The interface is as follows, and it looks quite feature rich: The built-in offline ChatGPT, when asked what generation it is, it says it is the third generation. I asked about the recently popular version number issue, and the default is still the comparison version number. I told it that it can still reflect correctly after switching to money. In addition to chatting, there are also functions for batch processing of images, such as lossless enlargement, watermark removal, coloring and photo restoration, and 3D conversion. It adopts batch processing mode for images, which requires you to first put the images to be processed in a file, and then configure the parameters. Like the image enlargement function, below are some blurry small emojis that retain the "emoji flavor". You can make them clear in batches. To remove watermarks, you can choose a position mask, but the effect can only be described as "usable". For more complex grid backgrounds, it is easy to see the smudging feeling. The rest is 3D image production, which is not about converting images into 3D models, but about creating 3D effects and parallel eye effects. There are also functions such as color restoration for images and videos, which feel like adding a layer of filter, but the effect falls far short of expectations for AI. To increase the resolution of a video, it will first be decomposed into images and processed one by one. If the hardware cannot keep up, this process can be time-consuming. In addition, this tool also comes with a small feature that decrypts and synthesizes videos cached by Bilibili, which can be used on both Android and Windows devices. After a simple experience, it feels like a mediocre little tool. The software author said that more AI models and features will be added in the future. If you can choose what functions you need and then package and download them locally, everyone's willingness to try will greatly increase. --- *Originally published on [wangmingqing](https://paragraph.com/@wangmingqing/software-experience)*