Summary
Fourteen years ago, Google launched an app called “Androidify.” It allowed you to customize a 2D Android robot character version of yourself. Over the years, Google has updated and removed Androidify, but it’s back again, and this time it’s using Gemini AI.
After sitting dormant for a long time, Google pulled the Androidify app from the Play Store in 2020. Then, last year,it was brought backwith a 3D robot and new customization options. At GoogleAII/O 2025, the company announced another big update to Androidify. This time, instead of manual customization, you give it a photo, and AI does the rest.
The process is extremely simple and straightforward. First, you have the option to snap a fresh new photo or grab one from your photo gallery. It doesn’t matter if you use a selfie or a full-body photo. Once you have a photo, you can choose the bot color. Then it’s just a matter of waiting for the output. In my experience, it usually takes around a minute or so to generate the image.
As with all things AI-related, the results are a mixed bag. Sometimes, it’s absolutely spot on, and other times it’s not quite right. Here are a couple of examples of times when it did really well.
I was really impressed with those results. The first image matches my hair, shirt, and pants pretty flawlessly. In the second image, I wanted to see if it could do multiple people, and not only did it do that, but it basically nailed both of our outfits. The detail in the last image was maybe the most impressive. It generated the correct color for the graduation stole and cap tassel.
The next batch of results is mostly fine, but there’s a little weirdness. The first image does an awesome job of recreating my blue and yellow tie-dyed shirt, but my beard is entirelybelowthe robot’s neckline. This happened pretty often. I challenged Gemini with a big group photo full of props in the last image, and it honestly performed okay. However, those are not robots—they look more like Barbie dolls—and only one is green. Seems like Gemini got overwhelmed and lost the plot.
So, for the most part, it’s pretty accurate, and more importantly, lots of fun to play around with. I do miss the customization of previous versions of Androidify, though. It was cool to be able to change your clothes and put stuff in your hands. There was more creativity involved, whereas now it’s basically limited to whatever you’ve worn in the real world. There is one big missing feature, though. I tried to use a photo of my dog, and the app said animals are not supported. Bummer.
Of course, Androidify is just a small taste of all the Gemini AI features that were announced at Google I/O. In a similar vein, Google is also using Gemini to generate images of what you look like wearing clothes from Google Shopping. The company also demoedAndroid XR glasseswith Gemini built in, launched a$250/month “Ultra” Gemini subscription,phone apps for NotebookLM, and anAI detector.
Ultimately, this is yet another AI tool that’s fun to play with, but lacks a certain quality of previous non-AI iterations. Sure, it’s easier, but you’re giving up some creative control to get there. At the end of the day, isn’t personalization the point of creating apersonalizedrobot version of yourself? AI tools allow us to offload work, but sometimes the journey is more important than the destination.
The new Androidify app is currently onlyavailable on Githuband needs to be compiled in order to actually install and use it. We’re not sure if Google will ever make it publically available on the Play Store.
