ChatGPT now adds the GIF feature, allowing users to create and edit animations with simple commands.

The new update was added by OpenAI with the Images 2.6 model on November 4, allowing users to create new, then adjust the GIF's animation speed, edit each frame and share with others. Besides, users can also download still images, line sketches... add commands to get animations as required.

According to OpenAI, users worldwide currently create 4 billion images per week through the ChatGPT, Codex and API platforms (application programming interface), equivalent to more than half a billion images created per day, higher than the 3 billion images per week announced in September. On the blog, the company said GPT-Image-2.6 creates images 50% faster than the previous version.

Only on the API, developers can choose two versions: Flare and Sunburst. Flare prioritizes speed, can create images in 6 seconds at low quality, and improves poster layouts and product photos. Sunburst focuses on detail quality and maintains better image consistency across medium or high quality edits.

Compared to previous versions, Images 2.6 also expands developer control during image creation and editing. The API adds a markup feature to pinpoint exactly where changes need to be made, and supports SVG export for websites, applications, and publications. Additionally, the complexity parameter allows adjusting the level of planning of the model, while the quality setting determines the detail and visual fidelity.

The ability to create images is currently free on ChatGPT. However, to use advanced features, users need to upgrade to higher packages. With the API, Images 2.6 Flare costs $8 per million tokens for input and $30 for output. A token is a basic unit of input data consisting of a word, part of a word, character, or punctuation that an AI model processes to produce an output.

According to The Verge, OpenAI's addition of the ability to create live GIF images shows a new direction for chatbots to expand from a static image creation tool to produce moving visual content. The new feature also surpasses most competitors, which do not support creating this type of photo, or have to go through the process of combining multiple still images.

The GIF (Graphics Interchange Format) format was born in 1987, developed by CompuServe to display and share color images on online services when Internet speed was limited. In 1989, CompuServe introduced version GIF89a, which added the ability to display animations, set transition times between frames and transparent backgrounds - the foundation for today's popular animated GIF format.

According to National Archives, GIFs are essentially created from a series of still frames that play consecutively, usually without sound, and can loop continuously. Thanks to their relatively small size, ease of sharing, and ability to convey short movements, GIFs are popular on social networks, messaging apps, websites, and advertising campaigns. This format is often used to create memes, illustrate operations, introduce products or make visual content more vivid.

Bao Lam summary

• Innovators struggle to prove 'humanity' in the age of AI
• OpenAI launched more than 20 AI products and features in one day
• OpenAI hires people to read users' ChatGPT conversations
• ChatGPT has a version for teenagers