TLDRocket
Sign in

The new ChatGPT Images is here

OpenAI

OpenAI just rolled out a new ChatGPT Images, and it's noticeably faster and sharper. Edits stick this time, and speed is up 4x, which actually changes how you use it.

Based on reporting by OpenAI — read the original for the full story.

Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error

OpenAI pushed out a refreshed ChatGPT Images today, and the headline number is speed: image generation is now up to four times faster than before. That's not a marginal tweak. Waiting ten or fifteen seconds for a render versus a couple seconds changes how people actually use the tool, turning it from a slow one-shot generator into something closer to an iterative sketchpad.

The bigger deal, though, is precision. OpenAI says the underlying model now handles edits more accurately and keeps details consistent across revisions, which has been a genuine pain point since image generation inside ChatGPT launched. Anyone who's tried to nudge a single element in a generated image while keeping everything else intact knows how often the whole picture would shift or degrade instead. If this update actually fixes that, it matters more than the speed bump.

This new model isn't confined to the ChatGPT interface. It's shipping simultaneously in the API as GPT-Image-1.5, meaning developers building their own apps and products get access to the same upgraded generation and editing capability, not a watered-down version. OpenAI is rolling the update out to all ChatGPT users starting today, so there's no separate tier or waitlist gating this one.

What's notable is the timing and framing: OpenAI is positioning this squarely as its flagship image model, not a side experiment. Image generation has become a competitive front with Google, Midjourney, and others iterating fast, and shaving generation time while improving edit fidelity is a direct response to where users have been complaining loudest.

My take — AI-written commentary, not fact-checked reporting

Faster and more consistent edits sound modest until you remember how much of AI image generation has been trial-and-error luck rather than actual control. If OpenAI genuinely nailed reliable editing, that's the unlock that matters more than any benchmark, because it's the difference between a toy and a tool people build workflows around.

Read more about this at: OpenAI

Related stories

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.