Google Taught Gemini to Make GIFs: 3 Seconds, 16:9 or 9:16 Only, No Loop
Google has added GIF generation to Gemini from text, images or video, but there is almost no control over it. The animation lasts about 3 seconds, and the frame is either landscape 16:9 or portrait 9:16. There is no square option. You can't pick the duration or resolution, or make a loop where the last frame matches the first. It's fine for a quick chat reaction, but branded stickers need a different generator.
AI-processed from TG: @cgevent (Метаверсище и ИИще); edited by Hamidun News
Imagine you want to quickly make an animated reaction for a chat or a short interface demo for a README. Google has built exactly that into the Gemini app: native GIF generation, meaning a short looping animated image, from text, images or video. But its capabilities are modest.
The clip lasts about 3 seconds, and the frame is either landscape 16:9 (like a TV screen) or portrait 9:16 (like a story), with no square format. You can't choose the duration or resolution, and there will be no "loop" where the last frame matches the first. According to the author of the Telegram channel "Metaversishche i IIshche", in terms of control over the result this is so far a weak answer to GIF generation in GPT Image 2.
6.
What Gemini Can Actually Do
GIF generation in Gemini works in three modes: from a text description, from an image, and from a video. You can feed it up to five images or one video. That is convenient: for example, you can upload several frames or a ready-made short clip and ask Gemini to turn it into an animation.
According to the author of the review, the model under the hood is Omni Flash 1.2. Google, however, does not publish detailed technical parameters. The exact frame resolution, the number of frames and the so-called pixel budget, meaning the total area limit across all frames of the animation, are unknown. The only known figure is the approximate duration: about 3 seconds.
Where the Limitations Start
The main complaint is about format. Gemini produces only landscape 16:9 or portrait 9:16 animations. You can't get a square GIF, which is still very popular in messengers and social networks.
There are a few more things the user does not control:
- duration: the clip is about 3 seconds, and that can't be changed;
- resolution: you can't choose the frame size;
- looping: the last frame is not matched to the first, so there will be no "seamless loop".
The last point is especially important. A good GIF usually relies on a loop: the motion ends where it began, and the eye doesn't notice the seam. If the seam is visible, the animation looks like a trimmed video fragment rather than a real GIF. For memes and quick reactions that is tolerable, but not for branded stickers, animated icons and banners.
Comparison With GPT and What Practitioners Should Do
The channel's author compares Gemini with GIF generation in GPT Image 2.6. For that one, he says, technical details about limits on total frame area have already appeared, and he rates the result much higher: "after GPT gifs this is just garbage". This is the subjective opinion of one reviewer, not the result of an independent test, and the piece contains no numerical quality comparison. But the signal is clear: in parameter transparency and control over the result, Google is so far behind.
What this means in practice:
- If you need a quick reaction or a joke animation for personal messaging, Gemini will do: it is available in the regular app and requires no API or setup.
- If you are building a product where a GIF must be predictable in size, duration and format, you can't rely on Gemini's feature: the parameters can't be set and are not documented.
- If a loop matters for your use case, you will have to either look for another generator or fix the result manually in a video editor.
- Before integrating, it is worth waiting for official documentation: it is not yet known whether API access will appear or whether the limits will be disclosed.
Honest Caveats
The entire review is based on the observations of the Telegram channel's author and the app's user interface. The source contains no official technical details from Google, so it can't be claimed that the limitations will remain forever. Features like this in apps often start as a stripped-down version and gain settings over time: choice of duration, aspect ratio, looping.
For now the picture is this: Google is catching up with competitors in the set of modes (text, image, video) but is far behind in controllability. For the industry, it is one more example of how the race to say "we have it too" finishes before the feature is brought to a level fit for real work.
- Nov 21, 2026Startup Osmo Trains AI on Smells Using 10 Million Human Ratings and Generates Fragrances From Text
- Nov 21, 2026North Korea Shows First Chatbot "Source of Wisdom": Science Only, No Politics
- Nov 21, 2026OpenAI Opens ChatGPT Spaces to Plus Subscribers: Shared Documents with AI
- Nov 21, 2026Instinct, valued at $10 billion, leads the market for AI agents you text like a messaging app
Key Questions Answered
What Gemini Can Actually Do?
GIF generation in Gemini works in three modes: from a text description, from an image, and from a video. You can feed it up to five images or one video. That is convenient: for example, you can upload several frames or a ready-made short clip and ask Gemini to turn it into an animation. According to the author of the review, the model under the hood is Omni Flash 1.2. Google, however, does not publish detailed technical parameters.
Where the Limitations Start?
The main complaint is about format. Gemini produces only landscape 16:9 or portrait 9:16 animations. You can't get a square GIF, which is still very popular in messengers and social networks. There are a few more things the user does not control: duration: the clip is about 3 seconds, and that can't be changed; resolution: you can't choose the frame size; looping: the last frame is not matched to the first, so there will be no "seamless loop". The last point is especially important.
What are the key details on Comparison With GPT and What Practitioners Should Do?
The channel's author compares Gemini with GIF generation in GPT Image 2.6. For that one, he says, technical details about limits on total frame area have already appeared, and he rates the result much higher: "after GPT gifs this is just garbage". This is the subjective opinion of one reviewer, not the result of an independent test, and the piece contains no numerical quality comparison. But the signal is clear: in parameter transparency and control over the result, Google is so far behind.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.