ChatGPT Gets Animated GIFs and Photo‑Memory Tricks with Images 2.6

OpenAI surprised a lot of us on November 4 by unveiling Images 2.6, the latest version of its image model that lives inside ChatGPT, the Codex coding helper and the public API. The headline‑grabber? Images now pop out about half as fast as before, and two brand‑new tricks go beyond static pictures.

First, you can turn a simple prompt or an uploaded photo into an animated GIF. Want a bouncing ball, a waving flag, or even a quick reaction meme? Just describe what should move, or feed ChatGPT a still image, and the system will spin a looping clip. After the GIF is generated you can tweak the playback speed, download it, or share it straight from the chat. At the moment the feature is still rolling out – some accounts see it today, others may need a day or two – and OpenAI hasn’t published any hard limits on how many GIFs you can make.

Second, the update adds a personal‑photo memory. Upload a picture, give it a nickname (say, @Pepper for your dog), and later summon it with an @‑mention in any conversation. The model can then treat that stored image as a reference – for example, “paint a watercolor portrait of @Pepper” – without you having to re‑upload the file. The saved photo lives in your account indefinitely, though the help pages are still vague about where you can manage or delete these references.

Generated images, on the other hand, are automatically filed under the “Images” tab of the conversation and disappear when you delete that chat. OpenAI does tag every AI‑created visual with a Content Credential and an invisible SynthID watermark, but GIFs don’t get a separate label yet. That raises a privacy question: if you store a real person’s photo and later animate it, the resulting GIF could place them in fictitious scenes. As of now OpenAI hasn’t spelled out safeguards against that kind of misuse.

The practical upside is clear, though. Faster generation means less waiting when you’re tweaking UI mock‑ups or brainstorming visual ideas. The GIF tool opens up quick meme‑making or simple motion‑design without needing a separate editor. And the photo‑memory feature is a handy shortcut for recurring subjects – think product shots, mascot images, or personal avatars.

All of these perks are available on every plan, including the free tier, which is a nice surprise for casual users. If you’re worried about your uploads being used to train future models, you can toggle that in the Data Controls section of the settings.

In short, Images 2.6 feels like a modest but genuinely useful upgrade. It won’t replace a full‑blown animation suite, but for quick loops and personal reference images, ChatGPT just got a lot more handy.