Gemini can now make GIFs, and they're free in Germany
Google has added a GIF generator to its Gemini app. Describe what you want in text, and it makes looping animations and funny memes. It runs on the Gemini Omni 1.2 Flash video model, which launched in early November. The feature is free in Germany and seven other countries.

What's new
Google has given its Gemini AI app a new trick: it can now create animated GIFs. These are the short, endlessly looping animations you know from chats and social media. According to the Chinese tech site ITHome, the feature works for seamless loop animations and for funny memes.
It runs on Gemini Omni 1.2 Flash, the video model Google introduced in early November. GIF creation is available to all Google AI subscribers, and it's also free in the US, Japan, South Korea, Australia, Canada, the UK, Germany and France.
ITHome says there are two ways to use it. You can type your request straight into the regular chat. Or you can pick the "GIF" option in the image and video tools and describe what you want to see. After a short wait, your animation is ready.
The model behind it: Gemini Omni 1.2 Flash
Gemini Omni is Google's family of AI video generation models. According to Crypto Briefing, the series debuted on May 19, 2026, at Google's I/O developer conference. Version 1.1 Flash added scene extension and first-and-last-frame control: you decide how a clip starts and ends, and the model fills in the motion between them. Version 1.2 Flash was announced on November 5, 2026, and is generally available to developers and businesses.
Here are the key upgrades, according to Android Authority, Pulse 2 and Crypto Briefing:
- Longer clips: Clips can now run up to 15 seconds instead of 10, a 50 percent increase. You can set the length to the exact second, anywhere from 3 to 15 seconds.
- Conversational editing: You can change a finished clip over several messages without regenerating it. You might swap the background, product, lighting, weather or aspect ratio. Google says anything you don't ask to change stays the same.
- Better reference fidelity: You can supply character images, product photos, style references, short motion clips of up to three seconds, and audio cues. Crypto Briefing reports 7 to 10 reference images per generation. Faces, clothing, packaging, text and brand colors are supposed to stay consistent across the whole video.
- More natural speech: Google says dialogue sounds smoother, lip sync is tighter, and multiple speakers have distinct voices.
- Camera moves: Complex continuous shots should come out with fewer unwanted cuts. Examples include FPV flights (a drone's-eye view), dramatic pull-back zooms and orbiting camera moves.
In Google's own words:
Characters, objects and scenes stay more consistent from the first frame to the last, with less visual drift or warping during motion and camera transitions.
Resolution, pricing and watermarks
Pulse 2 notes you can save money by drafting in 360p or 720p and upscaling only the takes you like to 1080p or 4K. Videos come with synchronized audio rather than silent. Pulse 2 also reports prices 30 percent lower than before. According to Crypto Briefing, 720p output via the API costs between $0.02 and $0.10 per second, depending on the source. That would put a full 15-second clip at roughly $0.30 to $1.50. Note that this range is only reported. Google hasn't confirmed it.
Everything the model generates carries SynthID, a digital watermark that people can't see. It's meant to let AI-generated content be identified later.
Where you can use Omni 1.2 Flash
Access varies by product:
| Platform | Access |
|---|---|
| Gemini app (15-second videos) | Google AI Ultra |
| Gemini app (GIFs) | All Google AI subscribers; free in Germany and other countries |
| Google Vids | Up to three 15-second videos per month for free |
| Flow | Free users too, limits unclear |
| Gemini API / Google AI Studio | Developers |
| Google Cloud | Enterprises |
| Runware, Segmind | Third-party providers |
Google Ads will use the technology too, and so will YouTube Shorts, according to Crypto Briefing. Early partners include:
- Adobe, which plans to build the model into Firefly, its image and video software
- Monks, a creative agency
- Havaianas, the footwear brand
- Whering, a fashion startup that is testing virtual try-ons
In one demo, Google created a 15-second virtual try-on from a single model photo and product shots of several outfits.
Also new: Gemini 4 Neon
ITHome also mentions a second Google model. Gemini 4 Neon launched on November 18 and targets complex coding, scientific research and enterprise workflows. Google says it strikes a better balance between reasoning, speed and cost.
Via the API it costs:
- $3 per million input tokens
- $15 per million output tokens
- $0.15 per million cached input tokens
Tokens are the word and syllable fragments AI models use to process text. Caching means text you send repeatedly is billed at a lower rate. Neon has four thinking levels: Low, Medium, High and Xhigh. In the Gemini app, it's available to Google AI Pro and Ultra subscribers.
The big picture
GIFs look like a toy at first glance, but they're a smart move. Google is bringing its expensive video model to a lot of people in a small, everyday form. And unlike many new AI features, this one is free in Germany from day one. If you want to make your own reaction GIFs or memes, you can try it right now. One video title among the sources for this story suggests ChatGPT now offers a GIF feature too. The race for fun features like this has begun.
The real substance is the model underneath. With 15-second clips, conversational editing and better consistency, Omni 1.2 Flash is clearly built for advertising, e-commerce and marketing. Google itself names some important limits. The model still struggles with fast sports scenes, abrupt cuts in scripted scenes and complex physical interactions.
A few questions remain open:
- There are no independent tests of the GIF feature's quality yet.
- It's unclear how many GIFs free users can make.
- SynthID labeling makes sense, but it only helps if platforms and users actually check for the watermark.
Sources
- ITHome (CN) ithome.com
- androidauthority.com androidauthority.com
- pulse2.com pulse2.com
- cryptobriefing.com cryptobriefing.com
- Bing News (en-US) neowin.net
- Google News (IL) news.google.com
- Bing News (en-US) letsdatascience.com
- Bing News (en-US) digitaltrends.com
Ask the AI assistant
Questions about this article, installation or background? Our assistant knows this page and runs on our own machine, not in someone else's cloud.
The Assistant is an AI (Gemma 4) and can make mistakes – please check important details against the source. Questions are used only to generate an answer and are deleted after one hour. Please don't enter any personal data.
Comments
What do you think? Already tried it? Write it here, no sign-up needed.
Comments are public. Links aren't allowed, and we delete abusive posts. You can delete your own comment at any time in this browser. Learn more in the Privacy Policy.