According to OpenAI, the new ChatGPT Images can follow instructions more closely and make more accurate modifications.
An improved version of OpenAI's visual generating tool, ChatGPT Images, has been released. The new tool offers significant gains in instruction-following, editing accuracy, and production speed thanks to its flagship picture generating model, GPT picture 1.5. Image generation and iterations are now up to four times faster than before, according to the San Francisco-based artificial intelligence (AI) behemoth. It's interesting to note that these enhancements put ChatGPT Images on par with Google's recently introduced Nano Banana Pro.
Features and Capabilities of ChatGPT Images
OpenAI described the updated ChatGPT Images feature in a post. The company has improved the tool's responsiveness to instructions with this version. Whether they are developing graphics from scratch or using pre-existing photos, users can define what they want, and the tool will produce outcomes that more closely match those directions. Additionally, the accuracy helps with editing work where undesired modifications can now be minimized.
Speed is one of the main features of the new ChatGPT Images. According to OpenAI, the revised model can create and modify images up to four times more quickly than it did previously. When individuals are refining a photo edit or testing variants on a design, for example, this can be quite important because shorter wait times result in a more effective process.
These features are underpinned by a new flagship picture generating model that takes the place of earlier image generation bases. According to OpenAI, it can now maintain more abstract characteristics, including lighting, composition, and people's looks, consistent across several revisions and can now more accurately follow human intent.
Additionally, ChatGPT now includes a dedicated Images interface thanks to OpenAI. This gives users a clearer area to explore visual concepts and distinguishes image generating from ordinary chat interactions. Both the website and the app's sidebar provide access to the area. Instead of typing lengthy and extremely comprehensive commands, users simply tap a variety of preset filters, upload an image, and make an adjustment in that manner. According to the company, the area serves as a place for inspiration.
All users worldwide, even those on the free tier, can now access ChatGPT Images. Additionally, by using GPT Image 1.5 in the application programming interface (API), developers can choose the AI model. Notably, a custom GPT will be offered for the previous iteration of the picture tool.