ChatGPT’s new image tool generates pictures four times faster than before

OpenAI has released an updated version of its image generation feature in ChatGPT. The new model creates images up to four times faster than the previous version and offers more precise editing capabilities. The company also introduced a dedicated section for image creation within ChatGPT’s interface.

The updated model, available through ChatGPT for all users and as GPT Image 1.5 through the API, focuses on maintaining consistency during edits. According to OpenAI, the system can change specific elements while preserving details like lighting, composition, and people’s appearance across multiple editing steps.

The model handles different types of edits, including adding, removing, combining, and transforming elements. OpenAI demonstrates this with examples ranging from practical photo edits to creative transformations like movie posters and style filters. Users can request changes through text prompts or select from preset options in the new Images section.

Text rendering has improved significantly. The model can now handle denser and smaller text, as shown in examples of infographics and detailed layouts. OpenAI states that instruction following has also become more reliable, enabling more complex compositions where relationships between elements remain intact as specified.

The new creation space in ChatGPT’s sidebar offers preset filters and trending prompts. The feature includes a one-time likeness upload option, allowing users to include their appearance in future creations without repeatedly uploading photos. Images continue to generate while users work on new requests, reducing waiting time.

OpenAI tested the new model against examples from its initial image generation launch. The company acknowledges that while the model shows clear improvements, results remain imperfect.

Bloomberg notes that the update comes as OpenAI faces growing competition from companies like Google, which recently introduced its Gemini 3 model and offers an image generator called Nano Banana. OpenAI CEO Sam Altman declared a “code red” situation in an internal memo following the Gemini 3 launch.

Sources: OpenAI, Bloomberg

About the author

Related posts:

Stay up-to-date:

Advertisement