OpenAI launches faster, more precise ChatGPT Images model for all users and API as GPT-Image-1.5

OpenAI has launched a major upgrade to its image generation stack with “ChatGPT Images” powered by the new GPT‑Image‑1.5 model, bringing faster generation, more precise edits, better text rendering, and cheaper pricing for both ChatGPT users and API developers.​

What is GPT‑Image‑1.5?

GPT‑Image‑1.5 is OpenAI’s latest image generation model that replaces GPT‑Image‑1 as the engine behind the new ChatGPT Images experience and the image generation API. It is designed for high‑quality visual creation and precise editing, with capabilities optimized for marketing, product visuals, design workflows, and creative content at scale.​

Key improvements and speed

The new model can generate images up to around 4x faster than the previous generation in common use cases, significantly reducing turnaround time for iterative creative workflows. Along with speed, GPT‑Image‑1.5 follows complex instructions more reliably, preserves fine details, and renders in‑image text (like logos, labels, or UI copy) more accurately than earlier versions.​

Advanced editing and control

A standout feature of GPT‑Image‑1.5 is its editing precision: you can upload an existing image and ask for targeted changes such as adjusting lighting, changing hairstyles or outfits, or adding a logo, while keeping identity, composition, and style consistent across multiple edits. This makes it especially useful for product photography tweaks, branding variations, and A/B creative testing where only specific elements should change while the rest of the scene remains stable.​

Availability in ChatGPT and API

For end users, the new ChatGPT Images experience appears as a dedicated “Images” entry point with preset styles and templates, allowing anyone to generate or edit visuals directly inside ChatGPT’s interface. For developers, GPT‑Image‑1.5 is accessible as a model name in the image generation and editing endpoints of the OpenAI API, making it straightforward to plug into apps, design tools, and automation workflows.​

Pricing and cost efficiency

OpenAI has reduced the cost of image inputs and outputs by about 20% compared with GPT‑Image‑1, so developers can generate and iterate on more images for the same budget. This stacks on top of the existing tiered pricing for image quality in the OpenAI platform, where low, medium, and high‑quality image outputs are billed at different per‑image rates to match use cases ranging from quick concepts to print‑grade artwork.​

Safety, rights, and enterprise focus

GPT‑Image‑1.5 ships with updated safety systems and content filters, as well as support for provenance metadata like C2PA, helping platforms signal that an image was AI‑generated and reducing the risk of misuse. OpenAI continues to state that users own the images they generate within the bounds of usage policies, which combined with higher fidelity and control positions GPT‑Image‑1.5 as a strong option for enterprise marketing, design, and product teams that need reliable, rights‑clear visuals at scale.