OpenAI adds GIF creation and faster image generation to ChatGPT

New capabilities for ChatGPT

OpenAI has released the Images 2.6 model, which is now integrated into both ChatGPT and Codex. This update introduces the ability to create GIFs, which are short looping animations, directly within the platform. Users can generate these animations by providing a text description of the scene they want to see.

The update is designed to be accessible to a wide audience, as it is available to users across all subscription tiers. This means whether you are a free user or a paid subscriber, you can access the new features on web, desktop, and mobile devices. The goal is to make motion content more accessible to everyone from hobbyists to professionals.

According to pcmag.com, the move into animation is a significant addition to the toolset. By allowing users to create moving images as easily as static ones, OpenAI is expanding the creative possibilities for what the platform can produce in a single prompt.

Speed and quality options

One of the primary technical improvements in Images 2.6 is the speed of the generation process. The company reports that the model produces photos 50% faster than the versions that came before it. This efficiency is particularly important for users who need to iterate quickly on multiple different concepts.

For developers and power users, the update provides two distinct choices for output: GPT-Image-2.6 Flare and Sunburst. These options allow users to decide whether they want to prioritize how quickly an image appears or how much detail and accuracy the final result contains.

The Flare model is built for high-speed results, capable of delivering outputs in as little as six seconds. In contrast, the Sunburst model is designed to produce more accurate images, especially when the user selects the medium or high quality settings. This distinction gives creators the flexibility to choose the right tool for their specific project needs.

Tools for creators and marketers

The addition of GIF support offers specific advantages for people working in marketing and social media. OpenAI suggests that these tools can be used to animate brand mascots for specific campaigns or to create reaction memes for social platforms. It also provides a way for creators to explore character movements using animated storyboards before they commit to full video production.

For those who already have established characters, the model allows for a more consistent workflow. Users can upload a rough sketch of a mascot or character and pair it with a text prompt to generate a GIF. This helps maintain a consistent visual identity while still utilizing the AI's ability to create motion.

The platform also includes features to refine the output after the initial generation. Creators can adjust the speed of the GIF or modify individual frames before they share the final product with others. These controls ensure that the final animation aligns with the specific timing and style required for the intended use case.

Reference images and memory

To streamline the creative process, the new update includes a way to save reference images. Users can save a character, a pet, or even a photo of themselves just once within the system. Once saved, these subjects can be called upon using their names in future prompts rather than requiring a new upload every time.

This feature helps solve a common problem in AI generation where it can be difficult to keep a subject looking the same across different images. By using a saved reference, the model can maintain consistency in the appearance of the subject while changing the actions or environments around them.

This functionality works alongside the new GIF capabilities to create a more cohesive experience for long-term projects. It allows for a more organized way to build out a visual library of assets that can be reused across various different types of content, from static posts to moving animations.

Scale of AI image usage

The release of Images 2.6 comes at a time of massive scale for OpenAI's image generation tools. The company shared that its platforms currently generate an average of 4 billion images every week. This high volume of usage highlights how deeply integrated these tools have become into the daily workflows of people around the world.

The move toward faster generation and GIF support is a direct response to this high demand. As more people use the tools for work and play, providing faster results and more versatile formats like GIFs becomes a priority for the developers. It helps the platform keep up with the pace of content creation in a digital environment.

While the technical details of the model are complex, the user experience is being simplified to focus on ease of use. By providing clear choices between speed and quality, and adding motion features, OpenAI is attempting to make the tool more useful for a broader range of creative tasks without increasing the difficulty of the prompts.