OpenAI is reportedly gearing up to launch a new AI image generation tool in the coming weeks, according to a scoop from The Information.
https://twitter.com/kimmonismus/status/2040338389526822933
A source familiar with the plans says the new model will be a significant leap forward, particularly in generating complex images and detailed diagrams. It's also expected to produce visuals that look more natural, moving away from the telltale "AI-generated" aesthetic that's become so familiar.
No specific release date was given, but a launch in the coming weeks would point to late April or May 2025.
What prompted the push
CEO Sam Altman reportedly lit a fire under his team to improve OpenAI's image tools after Google released its competing model, Nano Banana. The pressure to respond is a familiar dynamic in the increasingly crowded and competitive AI image generation space.
OpenAI already offers image generation through ChatGPT and its API, but this new model is described as a major step up in quality and capability. The company hasn't made any official announcement yet.
What the new model is expected to do
The big promise? Improved realism. The goal is to make it harder to spot that an image was AI-generated. It should also handle more intricate visual content, like detailed diagrams, which current models can sometimes fumble.
Pricing and platform details haven't been confirmed. The report is based on a single source, and OpenAI hasn't verified the timeline or feature set. The Information, which broke the story, is a tech publication known for its deep coverage of the AI industry.
This push comes as OpenAI continues to expand its product lineup across both text and image generation. Altman has been vocal in the past about keeping OpenAI's products competitive with Google and other rivals. This move follows a period of strategic shifts at the company, including the recent departure of key leaders like Sora lead Bill Peebles as OpenAI pivots toward enterprise and developer tools.