It seems OpenAI likes to boost its chatbot’s image-generation capabilities at a rapid pace.
Just two months ago, users were introduced to ChatGPT Images 2.5. That update brought faster photo generation, sharper image details, more precise editing, and improved tools for creating & sharing. And now that we’ve reached November, Images 2.5 has officially entered its 2.6 phase.
Images 2.6 also promises an even faster rate of photo generation. However, the most exciting aspect of this new update is being able to turn any prompt you can think of or any photo you upload into a GIF. Alongside that welcome feature is the ability to adjust the animation speed of that GIF, edit its individual frames and share the final result across your social media.
You’ll be happy to know that the latest ChatGPT Images update is available to all ChatGPT, ChatGPT Work, and Codex users across its desktop, mobile, and web apps. While you’re here, check out some of the fun prompts you can try out with ChatGPT Images 2.6 below.
ChatGPT Images 2.6 is here and it’s the ultimate GIF generator
I’ve already come across some fun GIF-making prompts, such as “Turn me into a tiny chibi version of myself doing an absurdly enthusiastic victory dance. Transparent background GIF” and “Create a seamless looping GIF of me in 90s anime style, with expressive eyes, flat cel-shading, and soft colors. Exaggerate my reaction, then ease back to the starting pose. Keep my identity and framing consistent. First ask, ‘How do you want to react?’ and offer emoji choices.”
I played around with the new Images 2.6 update myself by trying out some of the best prompts ChatGPT came up with. Feel free to give these a spin so you can turn yourself, your pet or anyone/anything else you can think of into a fun GIF (be sure to upload a photo of what you want ChatGPT to work with first, such as a selfie, a family picture or a piece of art you drew):
- Turn yourself into a reaction GIF: Turn me into a ’90s anime character in a seamless looping reaction GIF. Keep my recognizable facial features, hairstyle and outfit consistent. Start with a neutral expression, build to an exaggerated reaction, then return smoothly to the starting pose. Keep the camera fixed. Before generating, ask which reaction I want: laughing, shocked, unimpressed or celebrating.
- Give your pet a ridiculous victory dance: Create a seamless looping GIF of my pet doing an absurdly enthusiastic victory dance after earning a treat. Use a miniature clay-animation style, preserving their distinctive markings and proportions. Keep the background simple and the camera still. Make the final pose flow naturally into the first.
- Make your own animated sticker: Create a transparent-background GIF of me as a tiny chibi character celebrating a small victory. Give me an oversized head, recognizable facial features and an exaggerated fist pump. Keep the whole character visible throughout the animation, with no text or extra objects. Make it loop smoothly.
- Turn a doodle into a living character: Transform my doodle into a polished character while preserving its unusual shapes, proportions and personality. Create a seamless looping GIF in which it notices the viewer, waves enthusiastically and returns to its original pose. Use a handmade felt-puppet style, a plain background and a fixed camera. Keep all character details consistent.
- Build a personalized birthday GIF: Create a seamless looping birthday GIF featuring this person as a charming paper-cutout character. Preserve their recognizable features. Have them blow out candles on a small cake, followed by a burst of confetti, then transition smoothly back to the opening scene. Use warm colors, a fixed camera and no text.
Of course, you can tell ChatGPT to make changes to your newly generated GIF with prompts such as “Slow the animation down, keep the camera fixed and make the transition between the last and first frames smoother. Preserve the character’s appearance.”
Bottom line
It’s pretty wild to see just how quickly ChatGPT’s image generator has evolved over the last couple of months. Images 2.5 already impressed me with how much cleaner its AI-generated pictures looked (especially since it retained my features when it turned me into an action movie star and someone who would totally thrive in the 80s).
Now I’m excited to see how Images 2.6 takes my selfie and turns it into any kind of GIF I can cook up in my brain. I’m also going to push it to the limit to see what kind of creative GIFs it can deliver based on prompts that don’t have an uploaded photo attached to them.
Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok.
More from Tom’s Guide
- Nvidia's new AI laptop runs nearly 80% slower on Linux — but only when it's plugged in
- Meta's Muse AI can do TikTok dances now and it's coming to your Instagram Stories — here's how to use it
- I gave ChatGPT one mission ahead of GTA 6 — here's everything it told me to do to get ready for the biggest game of the year
