OpenAI Unveils Exciting Updates to ChatGPT, Enabling GIF Creation from Text and Images

OpenAI has launched the ChatGPT Image 2.6, enhancing the user experience with a set of new features. This updated model integrates seamlessly with ChatGPT, Codex, and APIs, marking a significant enhancement with the introduction of a function to create animated GIFs from both text and images. Users can now transform any ordinary photo, character, or product image into a looping GIF by providing just a few simple prompts. This allows for quick creation of animated reactions, memes, and storyboards suitable for social media.

The standout feature of this update is the ability to generate animated GIFs directly from photos. With Image 2.6, users can craft animations by using either a text prompt or an image. The new model provides the option to convert images of people, pets, and products into looping GIFs, making it easier to produce visually engaging content for social media posts. Users also have the capability to edit the finished GIFs according to their preferences.

Moreover, the update includes functionality for editing individual frames of the GIF and adjusting the speed of the animation. This enhancement grants users increased control over the final output compared to previous versions, allowing for a personalized touch before sharing the complete animated versions.

Additional Updates in ChatGPT

Alongside GIF creation, OpenAI has incorporated a feature allowing users to save and tag reference images within Image 2.6. This enables users to store pictures of people, pets, and characters for future use without the need to re-upload the same image every time a new project requires it. This easily accessible reference library should streamline the image creation process.

Furthermore, the image generation speed has increased by up to 50%, which significantly reduces the waiting time for users when creating and modifying images. Currently, the company reports that over four billion images are generated weekly across ChatGPT, Codex, and APIs, a considerable rise from three billion images recorded eight weeks ago during the launch of Image 2.5. On average, more than half a billion images are created daily.

New API Models for Developers

OpenAI has also introduced two new models within its API for developers: GPT-Image-2.6 Flair and Sunburst, each tailored to specific needs. The GPT-Image-2.6 Flair model is designed for fast image generation, capable of producing images in just six seconds at low-quality settings. This model aids in optimizing layout quality for tasks such as posters and product shots. In contrast, the Sunburst model is geared towards producing higher-quality, consistent results during editing sessions, particularly useful for images requiring attention to detail and uniform output.

Image Markup and SVG Support Features

Image 2.6 also includes an image markup option, previously available in ChatGPT, which allows users to flag specific areas of an image to indicate desired changes. Additionally, the new SVG export support makes it easier to use generated images for web, app, and print-related work. The company has also provided developers with a complexity control option, enabling them to determine how meticulously the model plans image generation. Low settings can be utilized for straightforward edits, while medium and high settings facilitate the creation of more complex scenes with multiple elements.

Importantly, developers have the flexibility to independently set the quality and complexity of images. This grants them greater control tailored to their specific project requirements.

Availability of ChatGPT Image 2.6

According to OpenAI, Image 2.6 is now accessible on ChatGPT, ChatGPT Work, and Codex platforms. Users across all subscription tiers on web, desktop, and mobile can take advantage of these features. Additionally, the GPT-Image-2.6 Flair and Sunburst models are available through the API, providing both general users and developers access to innovative image generation and editing tools.