💥Open AI Image Generation 2.0 is here - its officially #1💥

OpenAI rolled out ChatGPT Images 2.0 yesterday (April 21) and it's live for all of us today. Short rundown of what it is and where it fits into our work.

What changed:

Text inside images actually works now - image/text, posters, thumbnails, infographics all render cleanly.
Two modes: Instant (fast) and Thinking (reasons before it draws, keeps characters and objects consistent across frames).
Up to 8 images from one prompt with continuity - carousels, storyboards, and series hold together.
2K output, native 4K on paid plans, wide aspect ratio support, improved non-Latin text (Japanese, Korean, Hindi, Bengali, Chinese).
Paid plans include a full commercial license.

First reactions from the AI community:

The common take is that this is the moment AI image gen moves from "fun toy" to "asset production."

Text rendering and character consistency are the two things people keep pointing to as the real unlocks. There's also pushback - some designers are calling it a "slop renaissance" and flagging concerns about visual sameness across the web (I both agree and disagree, but also hugely emphasize with designers).

Where I see this opening things up for us:

Content creation - infographics, carousels, post designs, Reel covers, thumbnails, data visuals, illustrations, custom graphics, mockups, product photography. Basically every visual asset in your content stack.
Personal brand & storytelling - this is the layer I'm most excited about. How do you use it to tell your story, showcase your work, and make abstract ideas visual for the visual learners in your audience? Quote cards, announcement visuals, behind-the-scenes illustrations, concept explainers etc.
Educational content & services - workbooks, lesson slides, diagrams, process illustrations, course visuals. If you teach, this collapses a huge chunk of the "make it visual" work.
Print-on-demand & merch - huge! Designs, mockups, full merch drops, poster art. Paid plans include a commercial license so you can actually sell it.
Digital products - also huge. SVGs, stock images, digital artwork, illustration packs, bundles.
Client work - storyboards, mood boards, concept visuals, campaign mockups. A much faster way to pitch and deliver creative work.
Ad-creatives - creating multiple ad creatives based on your product, offer or content.

To be honest, like with everything that's released in the past year, your imagination is the limit.

I think of this as an asset creation tool that I will map into my services, products and personal brand/content creation activities.

📹Let me know if you want a tutorial and what use cases would you like me to experiment with

I have been playing around with it this morning and I have to admit - it's impressive. May be this will be OpenAI's moat towards Claude that will keep it afloat (together with Codex), because even Claude does not have any solid image generation in sight.

Some of my experiments below: 1. fed it my Claude Design tutorial (coming up today!) transcript and asked to create an infographic 🤯 2. carousel on the Open AI release in my brand colors - will be playing around with thumbnails today as well...

Open AI release with more examples - https://openai.com/index/introducing-chatgpt-images-2-0/

Promo video with examples: https://youtu.be/6SIDwCQZkU8?si=AU1mxpj69Qw7FtIM

12 comments