OpenAI Livestream: ChatGPT Images 2.0
by wahnfrieden on 4/21/2026, 6:50:00 PM
Comments
by: hahahacorn
One of the images in the blog (<a href="https://images.ctfassets.net/kftzwdyauwt9/4d5dizAOajLfAXkGZ74DJh/f3c89f754604050c0980a0bc9de2091c/images-2-cornell.png?w=3840&q=90&fm=webp" rel="nofollow">https://images.ctfassets.net/kftzwdyauwt9/4d5dizAOajLfAXkGZ7...</a>) is a carbon copy of an image from an article posted Mar 27, 2026 with credits given to an individual: <a href="https://www.cornellsun.com/article/2026/03/cornell-accepts-5-776-students-to-class-of-2030" rel="nofollow">https://www.cornellsun.com/article/2026/03/cornell-accepts-5...</a><p>Was this an oversight? Or did their new image generation model generate an image that was essentially a copy of an existing image?
4/21/2026, 8:13:29 PM
by: simonw
I've been trying out the new model like this:<p><pre><code> OPENAI_API_KEY="$(llm keys get openai)" \ uv run https://tools.simonwillison.net/python/openai_image.py \ -m gpt-image-2 \ "Do a where's Waldo style image but it's where is the raccoon holding a ham radio" </code></pre> Code here: <a href="https://github.com/simonw/tools/blob/main/python/openai_image.py" rel="nofollow">https://github.com/simonw/tools/blob/main/python/openai_imag...</a><p>Here's what I got from that prompt. I do not think it included a raccoon holding a ham radio (though the problem with Where's Waldo tests is that I don't have the patience to solve them for sure): <a href="https://gist.github.com/simonw/88eecc65698a725d8a9c1c918478ac41?permalink_comment_id=6112150#gistcomment-6112150" rel="nofollow">https://gist.github.com/simonw/88eecc65698a725d8a9c1c918478a...</a>
4/21/2026, 7:26:48 PM
by: minimaxir
HN submission for a direct link to the product announcement which for some reason is being penalized by the HN algorithm: <a href="https://news.ycombinator.com/item?id=47853000">https://news.ycombinator.com/item?id=47853000</a>
4/21/2026, 7:17:53 PM
by: simonw
Suggest renaming this to "OpenAI Livestream: ChatGPT Images 2.0"
4/21/2026, 7:43:22 PM
by: volkk
the guys presenting are probably all like 25x smarter than I am but good god, literally 0 on screen presence or personality.
4/21/2026, 7:53:25 PM
by: ____tom____
No mention of modifying existing images, which is more important than anything they mentioned.<p>I think we all know the feeling of getting an image that is ok, but needs a few modifications, and being absolutely unable to get the changes made.<p>It either keeps coming up with the same image, or gives you a completely new take on the image with fresh problems.<p>Anyone know if modification of existing images is any better?<p>Anything better that OpenAI?
4/21/2026, 7:58:23 PM
by: dktp
One interesting thing I found comparing OpenAI and Gemini image editing is - Gemini rejects anything involving a well known person. Anything. OpenAI is happy to edit and change every time I tried<p>I have a sideproject where I want to display standup comedies. I thought I could edit standup comedy posters with some AI to fit my design. Gemini straight up refuses to change any image of any standup comedy poster involving a well know human. OpenAI does not care and is happy to edit away
4/21/2026, 8:13:39 PM
by: bensyverson
I caught the last minute of this—was it just ChatGPT Images 2.0?
4/21/2026, 7:22:50 PM
by: amunozo
This is not as exciting as previous models were, but it is incredibly good. I am starting to think that expressing thoughts in words clearly is probably the most important and general skill of the future.
4/21/2026, 8:03:44 PM
by: Melatonic
Can it generate anything high resolution at increased cost and time? Or is it always restricted?
4/21/2026, 8:21:05 PM
by: esafak
<a href="https://openai.com/index/introducing-chatgpt-images-2-0/" rel="nofollow">https://openai.com/index/introducing-chatgpt-images-2-0/</a>
4/21/2026, 7:32:42 PM
by: bitnovus
great obfuscation idea - hidden message on a grain of rice
4/21/2026, 7:21:15 PM
by: bitnovus
No gpt-5.5
4/21/2026, 7:22:28 PM
by: szmarczak
Wow, the difference between AI and non-AI images collapses. I hate the future where I won't be able to tell the difference.
4/21/2026, 7:45:38 PM
by: rqa129
Can it generate Chibi figures to mask the oligarchy's true intentions on Twitter and make them more relatable?
4/21/2026, 7:38:18 PM
by: sho_hn
In 5 years and 3 months between DALL-E and Images 2.0 we've managed to progress from exuberant excitement to jaded indifference.
4/21/2026, 7:49:37 PM
by: brianbest101
[dead]
4/21/2026, 7:35:15 PM
by: zb3
Image generation? Hmm, would be cool if OpenAI also made a video-generation model someday..
4/21/2026, 7:24:22 PM
by: aliljet
I am hopeful that OpenAI will potentially offer clarity on their loss-leading subscription model. I'd prefer to know the real cost of a token from OpenAI as opposed to praying the venture-funded tokens will always be this cheap.
4/21/2026, 7:21:21 PM