Hey there, 

ChatGPT can now turn a photo of your pet into a looping GIF, and that’s just one of the things that shipped this week. Google rolled out two new Gemini models, and Anthropic released its most capable Claude model yet. We’ll go through what’s new, who can use each one, and how to try the GIF maker yourself.

ChatGPT Can Now Make GIFs

OpenAI launched ChatGPT Images 2.6 on Nov 4, and the biggest addition is GIFs. In ChatGPT, you can turn a prompt, sketch, or photo into a looping GIF, then adjust the speed or make edits. You can also save named reference images, like a pet or a character, and bring them back in later chats by typing @ and the name. GIF creation is still rolling out, so it may not appear on your account yet, and it can’t edit a GIF you upload. To try it, upload a photo of your pet and ask ChatGPT to make a looping GIF where they tilt their head and wag their tail.

Gemini Hydrogen Is Google’s New Everyday Model

Google released Gemini 4 Hydrogen on Nov 12 as its new everyday model, replacing Gemini 3.8 Flash. Google calls it a daily driver built for speed, and says it completes tasks about 2.5 times faster than the model it replaces while making over five times fewer fabricated claims. On the Artificial Analysis Intelligence Index, it scores 51 at its default setting, just ahead of the larger Gemini 4 Argon’s 50. It also powers Google Search and is in the Gemini app for everyone, though people on the unpaid plan get it through automatic model selection rather than picking it themselves. In my opinion, this is the one that matters most to regular people, because it’s quietly running behind the Google features you probably already use.

Gemini Carbon Is Now Rolling Out to Ultra Subscribers

We covered Gemini 4 Carbon on Oct 30, and it’s now rolling out more widely. As of Nov 10, Google AI Ultra subscribers can use it in the Gemini app, AI Mode in Search, and Antigravity, and paying developers and businesses can use it too. Google positions it as its strongest model, built for the hardest problems and long-running tasks. It debuted at #1 on the Artificial Analysis Intelligence Index, and Google says it makes about half as many unsupported claims as Argon, flagging missing evidence instead of guessing. Google’s own teams have also watched it work for days at a time on complex engineering problems. In other words, Carbon is the model Google is pointing at for your most demanding work.

Claude Fable 5.2 Is Built for Hard Technical Work

Anthropic released Claude Fable 5.2 on Nov 9 on all paid plans, and calls it its most capable model for technical work like coding, finance, and science. It’s also up to 30% faster than the previous version. Fable is the public version of Claude Mythos 5.2, which only vetted organizations can use, and it carries extra safeguards for cybersecurity and biology that Anthropic says now block about 55% fewer harmless requests than before.