Guide

ChatGPT Image Generator: How It Works in 2026

How the ChatGPT image generator works in 2026: DALL-E removed from the API on May 12, 2026, GPT Image models, plan limits, and when a dedicated tool wins.

ChatGPT Image Generator: How It Works in 2026

The ChatGPT image generator no longer rests on DALL-E. OpenAI's deprecation list shows dall-e-2 and dall-e-3 removed from its API on May 12, 2026, with gpt-image-2, gpt-image-1 and gpt-image-1-mini as the recommended replacements.

You still type a prompt into the same chat box. What follows covers how OpenAI says to use it, what its plans cost, where it fits, and when you should reach for something else. Prices were read on OpenAI's pricing page in October 2026: ChatGPT Go is $8 a month, Plus is $20 a month, and Pro starts at $100 a month.

What changed since DALL-E

DALL-E was OpenAI's earlier image model, and the company's API deprecations page lists its removal on May 12, 2026, after notifying developers on November 14, 2025. The replacements it recommends are the GPT Image models: gpt-image-2, gpt-image-1 and the smaller gpt-image-1-mini.

On the ChatGPT side, OpenAI's image generation documentation describes asking ChatGPT to generate or edit images in a chat, and names gpt-image-2 as the model behind built-in image generation in its Codex app, CLI and IDE extension. For programmatic use, OpenAI points to its Image generation API.

If you want the wider picture of how the chat side of the product works, our guide on how to use ChatGPT covers the fundamentals that make image prompting work better.

How to use it

You generate an image by describing it in plain language, then attach a reference image when you want ChatGPT to edit it or use it as visual guidance. OpenAI's documentation gives a short recipe for the prompt: explain the image's purpose or audience, name the subject and what is happening, describe the setting, composition and style, add framing, dimensions, lighting, colors or materials when they matter, and state what the image must not contain. It adds that a useful prompt is often only one to three clear sentences.

The real advantage over standalone tools is iteration. OpenAI's advice is to start with the core idea and then make small, targeted revisions, adjusting one element at a time so the composition does not drift. When you edit an existing image, say exactly what should change and what must stay the same. In the desktop app you can select a generated image, switch to a canvas view of the images from the same chat, and add comments to one or more images before sending your revisions.

A few things OpenAI's documentation covers directly:

  • Editing your own uploads. Attach a photo and describe the change you want, and say what must stay the same.
  • Multiple reference images. Identify each image by order and explain how they relate, for example one image for the content and another for the style.
  • Text in images. Keep in-image text short, put the exact words in quotation marks, and describe font style, size, color and placement.
  • Infographics and dense layouts. OpenAI says image generation can help draft posters, labeled diagrams and timelines, but for dense copy or production-critical typography you should review every word and finish the asset in a design tool when needed.

Free vs paid: what you actually get

OpenAI's pricing page lists limited image creation on Free, more image creation on Go, more complex and accurate image creation on Plus, and unlimited and faster image creation on Pro, subject to abuse guardrails. It prints no fixed daily image cap, and OpenAI's documentation says image availability and usage limits in ChatGPT on the web depend on your plan and workspace settings. So the honest answer to "how many images do I get" is to open ChatGPT and check what your own account offers.

The plans themselves are simple to compare. OpenAI lists Free at $0 a month, Go at $8 a month, Plus at $20 a month, and Pro from $100 a month, with Pro plans at $100, $200 and $500. For Codex, OpenAI says image generations use your included limits about 3 to 5 times faster on average than similar turns without an image, depending on image quality and size, and that image generation is not available on the Free plan in Codex.

For almost everyone reading this, the practical advice is: start free, and upgrade only once you hit the limits often enough to be annoyed. What you pay for is more usage, not a different prompt box.

Where it is strong

The generator earns its place for a specific profile of work. It fits when you want a usable image fast without leaving your chat, when you expect to revise it a few times, and when the image needs a short piece of text. Blog headers, social posts, quick product mockups, slide illustrations, thumbnail concepts, and first-draft visuals all fit here. It is also beginner-friendly because there is nothing to learn beyond describing what you see in your head.

Conversational editing is the quiet superpower. Being able to say "keep everything, just swap the season to winter" and get a targeted edit back, instead of re-rolling the whole image, saves more time than any single prompt trick.

Where it falls short

Being honest here matters more than being enthusiastic. OpenAI's own guidance tells you where to be careful: long or dense text needs a word-by-word review, and production-critical typography belongs in a design tool. For revisions, OpenAI recommends changing one element at a time so the composition and other important details do not drift.

Commercial use: what to check

Before you put a generated image on a client invoice, read OpenAI's terms of use and usage policies. The terms say that, as between you and OpenAI and to the extent permitted by applicable law, you own the output, and they warn that output may not be unique, because other users may receive similar output. OpenAI's image documentation adds three practical points: use likenesses with care and confirm you have permission when depicting a real person, ask for an original treatment instead of imitating a specific brand, product, artist or artwork, and know that credit to OpenAI is optional.

Whether a purely AI-generated image can be copyrighted depends on the law where you work, so if exclusivity matters, add meaningful human editing on top and ask a lawyer.

When a dedicated tool is the better call

ChatGPT is the flexible generalist. Midjourney, Ideogram and Adobe Firefly are specialist image tools worth comparing with it, and each prices and licenses its output differently, so read the vendor's current plan page and terms before you commit.

A useful setup for most creators is ChatGPT for fast, iterative work, plus one specialist for the jobs where quality or style control is non-negotiable. To compare the full field side by side, see our roundup of the best AI image generators, and if you are choosing which assistant to build your workflow around, our best AI chatbots comparison is worth a read.

What these tools actually cost

We price every tool we review, so this is measured rather than estimated. Across 429 tools, 293 publish a price and 33% offer a free tier. Among content creation tools, the median entry plan is $15 a month, which sits below the $24 median across every category we price.

The spread matters more than the median. Half of the content creation tools sit between $9 and $24, and the range runs from $5 to $99. A quoted "starting at" price near the bottom of that range usually means per-seat add-ons land on top of it.

Price point Content creation tools All tools
Cheapest paid plan $5 $1
Lower quartile $9 $10
Median $15 $24
Upper quartile $24 $49
Most expensive $99 $990
Tools measured 23 293
Content creation tools: what the entry plan costs Content creation lower quartile$9Content creation median$15Content creation upper quartile$24All tools median$24
Median advertised entry price/mo. Source: Dupple pricing index, 293 tools with public pricing out of 429 reviewed, 2026-08-19.

FAQ

Is ChatGPT still using DALL-E for images?

OpenAI removed dall-e-2 and dall-e-3 from its API on May 12, 2026, and recommends gpt-image-2, gpt-image-1 or gpt-image-1-mini instead. OpenAI's current image generation documentation for ChatGPT describes generating and editing images by asking in the chat and does not mention DALL-E. You use it the same way, by describing what you want.

Can I generate images with the free version of ChatGPT?

Yes, with limits. OpenAI's pricing page lists limited image creation on the Free plan and more on paid plans, and its documentation says image availability and usage limits in ChatGPT on the web depend on your plan and workspace settings. Paid plans are Go at $8 a month, Plus at $20 and Pro from $100, and upgrading makes sense only when the limits get in your way.

Can I use ChatGPT images commercially?

OpenAI's terms of use say that, as between you and OpenAI and to the extent permitted by law, you own the output, and its usage policies still apply. OpenAI's documentation also says credit is optional, asks you to confirm permission when depicting a real person, and recommends an original treatment rather than imitating a specific brand or artist.

Which is better for text in images, ChatGPT or a design tool?

OpenAI says to keep in-image text short and specify it precisely. For dense copy or production-critical typography it recommends reviewing every word and finishing the asset in a design tool. For logos and detailed posters, a specialist such as Ideogram may give you more layout control.

AIpresso covers the AI tools worth your time in one short read a day. Join the creators and marketers who start their morning with the updates that actually matter.

Some offers on this page may be paid placements or contain affiliate links.

AIpresso: your daily AI brief

Free daily newsletter, read in 5 minutes.

Subscribe free