OpenAI announces ChatGPT Images update with improved performance and consistency

ChatGPT

OpenAI has announced a major update to ChatGPT Images, introducing improvements focused on precision, speed, and reliability. Powered by a new flagship image generation model, the update is rolling out globally across ChatGPT for all users and through the OpenAI API as GPT Image 1.5. The update aims to make image creation and editing more aligned with user intent, whether the task involves small adjustments or complete visual transformations.

One of the key highlights of the update is improved editing accuracy. OpenAI says the new model can make targeted changes to images while preserving important details such as lighting, composition, facial likeness, and background elements. This allows users to edit specific objects, clothing, or styles without affecting unrelated parts of the image. The improvement is especially useful for photo edits, branding visuals, and iterative creative work where consistency across multiple edits is critical.

Performance is another major focus of the update. ChatGPT Images can now generate images up to four times faster than before, reducing wait times and enabling quicker experimentation. OpenAI has also introduced the ability to continue generating new images while previous ones are still processing. This improvement supports faster creative workflows, making the tool more suitable for professionals working on tight deadlines.

The updated model shows noticeable gains in instruction following. It is better at understanding detailed and structured prompts, maintaining relationships between multiple elements, and producing layouts that closely match user requests. OpenAI highlighted improvements in tasks such as grid layouts, object placement, and multi-step compositions, which previously posed challenges for image generation systems. This makes ChatGPT Images more dependable for precision-driven use cases rather than purely exploratory ones.

Text rendering has also improved significantly. The new model can handle smaller and denser text with better clarity and legibility. This enhancement is particularly relevant for creating posters, infographics, advertisements, and document-style visuals where readable text is essential. OpenAI notes that these improvements reduce common issues such as distorted characters or incorrect spacing.

Alongside the model update, OpenAI is introducing a dedicated Images space within ChatGPT. Available through the sidebar on chatgpt.com and the mobile app, this new experience includes preset styles, trending prompts, and creative filters. The goal is to make image creation more accessible and intuitive, allowing users to explore ideas without writing detailed prompts every time.

For developers and businesses, GPT Image 1.5 is now available through the OpenAI API. OpenAI says image inputs and outputs are around 20 percent cheaper compared to the previous generation, making it more practical for large-scale applications such as ecommerce catalogs, marketing assets, and branded visuals. The model is designed to preserve logos and key visuals more consistently across edits.

OpenAI has also mentioned the limitations of the new image generation model, including challenges with certain artistic styles, multilingual text, and scenes involving many small faces. However, the company views this update as a meaningful step toward more controllable and reliable image generation. OpenAI says further improvements are planned as image generation continues to evolve across creative and professional use cases.