I’ve been a little bit addicted to rendering out images of toys generated by ChatGPT recently. The images below are toys based on some explainer video characters we’ve created for our clients. In case you’re wondering, the way to achieve this is to go to ChatGPT through their app or via a browser, sign in and then upload a reference image. Once you’ve done that, you need to add a prompt telling ChatGPT that you want it to use the image and render it into whatever style of toy you desire. It’s generally a good idea to add in the instruction to render the image photorealisticly.