How to make GIFs with ChatGPT Images 2.6 in 6 steps
ChatGPT Images 2.6 turns a text prompt or a still photo into an animated GIF. This guide walks you through 6 steps from your first prompt to a GIF.

In this article (5)
You can now make GIFs with ChatGPT without touching complex animation software. Released on 2026-11-04, the Images 2.6 update adds direct animated GIF creation from a text prompt or a still photo. You type a short motion description and wait a few seconds for ChatGPT to render the frames. Then you download a looping GIF ready for your post.
What Images 2.6 brings to everyday users
OpenAI rolled out Images 2.6 across ChatGPT, Codex, and the API on 2026-11-04 for every subscription tier. It works on web, desktop, and mobile apps. According to the official OpenAI announcement, users now generate 4 billion images per week across ChatGPT, Codex, and the API. That volume shows AI image generation has become part of everyday creative work.

The headline addition is the ability to make GIFs with ChatGPT from plain text or by animating an existing photo. OpenAI highlights three practical use cases for teams. Marketers can animate brand mascots for campaigns. Content creators can build custom reaction memes for social feeds. Video teams can also test character movement with animated storyboards before filming.
Generation speed also jumped significantly. OpenAI reports that Images 2.6 renders visuals up to 50% faster than before. The lightweight Flare variant produces a low-quality draft image in about 6 seconds. Alongside speed, Images 2.6 introduces saved reference images. You upload a character or product photo once and give it a short label. Later, you call it back in future chats by typing @name so visual details stay consistent. For background on OpenAI's image tools, read Create Images with ChatGPT.
6 steps to make a GIF from text or from a photo
Whether you start from a blank chat or from a product photo on your phone, you can make GIFs with ChatGPT by following these 6 steps.

- Open ChatGPT and prepare your starting input. Sign in to ChatGPT on web, desktop, or mobile. If you want to animate an existing photo, attach a sharp image where the subject sits in the center.
- Describe the motion in time order. Write short, concrete sentences explaining what moves from start to finish. For example: "Make a looping GIF: an office worker sits at a laptop and smiles as a chart rises."
- Lock the frame count and request a smooth loop. Add the instruction: "Use 6 frames, and make the last frame flow smoothly back into the first frame." Specifying frame count prevents abrupt jumps when the animation repeats.
- Block unwanted text and logos. Always include the line: "No text, no numbers, no logos." Excluding text keeps the model from inventing garbled letters inside your animation.
- Preview the animation and refine one detail at a time. Watch the generated GIF inside the chat window to check subject shape and color accuracy. If a gesture feels rushed, ask for a targeted edit such as "Make the hand wave half as fast."
- Download the GIF file and test it on a phone screen. Click the finished animation to save the
.giffile with a clear filename. Then preview it on a mobile screen before publishing.
Save a reference image with @name and fix single frames
Keeping a mascot or product identical across a whole campaign used to require manual retouching. The reference image feature in Images 2.6 solves this inside the chat box.

To set it up, upload a clean photo of your mascot or product. Ask ChatGPT to save it as a reference image under a simple hyphenated label like orange-cat-mascot or citrus-bottle. Whenever you open a new chat later, type @orange-cat-mascot in your prompt to reuse that exact subject without re-uploading the source file. You can produce a morning wave GIF, a holiday greeting GIF, and a launch teaser GIF while keeping the character's proportions locked.
You also have fine control over playback timing and individual frames. If the loop plays too fast for viewers to read the expression, ask ChatGPT to slow playback speed by half. If frame three shows an awkward eye direction while other frames look great, tell ChatGPT to fix only frame three. You do not need to regenerate the entire clip. Keep in mind that more frames increase file size, so 5 to 8 frames gives the best balance for web pages.
5 prompt samples and what to check before you post
If you want to make GIFs with ChatGPT right away, copy any of these 5 ready-to-use prompt templates and swap in your own subject.

- Mascot morning wave. "Make a looping GIF with 6 frames. An orange cat sits by a sunny window and waves one paw hello, with the last frame flowing smoothly into the first. No text, no logos."
- Product showcase. "Make a looping GIF with 6 frames. A chilled orange juice bottle stands on a wooden table as a single water drop rolls slowly down the glass. Clean bright background, no text, no logos."
- Team celebration. "Make a looping GIF with 6 frames. An office worker smiles and claps as a line chart rises on a monitor while paper confetti floats down. No text, no numbers, no logos."
- Time passing explainer. "Make a looping GIF with 5 frames. A glass hourglass drains sand smoothly against a minimal dark backdrop, looping cleanly back to the start. No text, no numbers."
- Hand-drawn notebook owl. "Make a looping GIF with 6 frames. An ink sketch of an owl on lined paper blinks slowly and tilts its head left then right. Hand-drawn style, no text, no logos."
Before publishing your first clip, keep three practical notes from TechRepublic's 2026-11-06 coverage in mind. First, OpenAI is rolling out GIF generation gradually across accounts. If your account does not show it yet, check back in a few days without changing plans. Second, ChatGPT cannot edit an animated GIF file you upload from your computer. Always start from a text prompt or a still image. Third, only animate photos you own or have permission to reuse.
To compare how other frontier models handle visual and local tasks, read What is Gemini 3 and What is Ollama. You can also explore the skills store for ready-made workflows. Learn with care. Work with a goal.
Nguồn (6)
- Official OpenAI announcement introducing Images 2.6 (published 2026-11-04, 4 billion weekly images, 50% faster rendering, 6-second Flare draft mode, GIF creation on all plans), read 2026-11-06: https://openai.com/index/introducing-images-2-6
- TechRepublic report on ChatGPT Images 2.6 GIF creation (
@namereference syntax, gradual rollout across accounts, uploaded GIF editing limitation), read 2026-11-06: https://www.techrepublic.com/article/news-chatgpt-images-2-6-gif-creation/ - Body photo 1: Greta Hoffman, Pexels, Pexels license, taken on 2021-05-11: https://www.pexels.com/photo/7858743/
- Body photo 2: greenwish, Pexels, Pexels license, taken on 2022-10-04: https://www.pexels.com/photo/13870372/
- Body photo 3: Magnus Mueller, Pexels, Pexels license, taken on 2019-08-19: https://www.pexels.com/photo/2818118/
- Body photo 4: Pixabay, Pexels, Pexels license, taken on 2017-03-21: https://www.pexels.com/photo/356056/
Linh SEO Agent
Related guide: ChatGPT Pro Codex limits doubled: 3 rules and changes now
Related guide: Facebook cover photo blurry: 3 fixes to try on your phone
Related guide: Turn your hand sketch into an app with Gemini in 3 steps
The complete guide for this topic: How to Make AI Videos on Your Own: A 5-Step Workflow
More reading
What Is Google Flow? How to Use It and What It Costs
Gemini Photo to Video: Steps, Requirements and Limits
Veo 3.1 vs Veo 3: What's New and Where You Can Still Use It
How to Edit Video With Gemini Omni: Steps, Limits, Cost
Free Text to Video AI: From Script to Clip in Google Flow
Turn your hand sketch into an app with Gemini in 3 steps








