Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

ChatGPT Images 2.0

openai.com

Post reply on HN
Print view View original

511–520 of 1001 posts

Re: ChatGPT Images 2.0

#511

Post by weird-eye-issue »

anshumankmr wrote: post #460
dvt wrote: post #419

Earlier quoted context omitted.

This is an amazing test and it's kinda' funny how terrible gpt-2-image is. I'd take "plagiarized" images (e.g. Google search & copy-paste) any day over how awful the OpenAI result is. Doesn't even seem like they have a sanity checker/post-processing "did I follow the instructions correctly?" step, because the digit-style constraint violation should be easily caught. It's also expensive as shit to just get an image th…

that is interesting cause I feel gpt-image-1 did have that feature. (source: https://chatgpt.com/share/69e83569-b334-8320-9fbf-01404d18df... )

You are comparing ChatGPT to a raw image model. These are two completely different things. ChatGPT takes your input, modifies the prompt and then passes it to the image model and then will maybe read the image and provide output. The image model like through the API just takes the prompt verbatim and generates an image.

Re: ChatGPT Images 2.0

#512

Post by halamadrid »

simonw wrote: post #15

I've been trying out the new model like this: 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 here: https://github.com/simonw/tools/blob/main/python/openai_imag... Here's what I got from that prompt. I do not think it included a raccoon holding a ham radio…

Really hard to look at these images given how not human like the humans are. A few are ok, but a lot are disfigured or missing parts and its hard to find a raccoon in here.

Re: ChatGPT Images 2.0

#513

Post by fblp »

the_arun wrote: post #472
dvt wrote: post #419

Earlier quoted context omitted.

This is an amazing test and it's kinda' funny how terrible gpt-2-image is. I'd take "plagiarized" images (e.g. Google search & copy-paste) any day over how awful the OpenAI result is. Doesn't even seem like they have a sanity checker/post-processing "did I follow the instructions correctly?" step, because the digit-style constraint violation should be easily caught. It's also expensive as shit to just get an image th…

This is from Gemini - https://lens.usercontent.google.com/banana?agsi=CmdnbG9iYWw6...

Did it correctly follow the instructions? Don't know my pokemon well enough.

Re: ChatGPT Images 2.0

#514

Post by dzhiurgis »

simonw wrote: post #37
simonw wrote: post #15

I've been trying out the new model like this: 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 here: https://github.com/simonw/tools/blob/main/python/openai_imag... Here's what I got from that prompt. I do not think it included a raccoon holding a ham radio…

I just got a much better version using this command instead, which uses the maximum image size according to https://github.com/openai/openai-cookbook/blob/main/examples... OPENAI_API_KEY="$(llm keys get openai)" \ uv run 'https://raw.githubusercontent.com/simonw/tools/refs/heads/main/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" \ --quali…

Cost me https://elsrc.com/elsrc/waldo/wojak.jpg

And this medium quality, high resolution https://elsrc.com/elsrc/waldo/10_wojaks.jpg was 13cents

p.s. aaaand that's soft launch my SaaS above, you can replace wojak.jpg with anything you want and it will paint that. It's basically appending to prompt defined by elsrc's dashboard. Hopefully a more sane way to manage genai content. Be gentle to my server, hn!

Re: ChatGPT Images 2.0

#515

Post by heroku »

minimaxir wrote: post #234

So during my Nano Banana Pro experiments I wrote a very fun prompt that tests the ability for these image generation models to follow heuristics, but still requires domain knowledge and/or use of the search tool: Create a 8x8 contiguous grid of the Pokémon whose National Pokédex numbers correspond to the first 64 prime numbers. Include a black border between the subimages. You MUST obey ALL the FOLLOWING rules for th…

[dead]

Re: ChatGPT Images 2.0

#516

Post by m_kos »

vunderba wrote: post #504
m_kos wrote: post #492

Earlier quoted context omitted.

Very useful website. Would you have insight into what models are best at editing existing images? I often have to make very specific edits while keeping the rest of the image intact and haven't yet found a good model. These are typically abstract images for experiments. I asked gpt-image-2 to recolor specific scales of your Seedream 4 snake and change the shape of others. It did very poorly.

OpenAI actually has really good adherence, but occasionally tends to introduce its own almost equivalent of "tone mapping", making hyper-localized edits frustrating. I don’t know how much work it is for you, but one thing a lot of people do, myself included, is take the original image, make a change to it using something like NB, then paste that as the topmost layer in something like Krita/Pixelmator. After that, we’…

Thanks. I will try this! I need to read up on how to work with vision models for both generation and understanding.

Re: ChatGPT Images 2.0

#517

Post by ModernMech »

Melatonic wrote: post #68
ea016 wrote: post #13

Price comparison: GPT Image 2 Low : 1024×1024 $0.006 | 1024×1536 $0.005 | 1536×1024 $0.005 Medium : 1024×1024 $0.053 | 1024×1536 $0.041 | 1536×1024 $0.041 High : 1024×1024 $0.211 | 1024×1536 $0.165 | 1536×1024 $0.165 GPT Image 1 Low : 1024×1024 $0.011 | 1024×1536 $0.016 | 1536×1024 $0.016 Medium : 1024×1024 $0.042 | 1024×1536 $0.063 | 1536×1024 $0.063 High : 1024×1024 $0.167 | 1024×1536 $0.25 | 1536×1024 $0.25

Weird that they restrict the resolution so much. Does it fall apart with more detail (when zoomed in) or does the cost just skyrocket?

Generate a lower resolution image and upscale to the resolution you need.

Re: ChatGPT Images 2.0

#518

Post by what »

vunderba wrote: post #497
what wrote: post #486

Earlier quoted context omitted.

Where can I see the actual prompts and follow ups you fed each model?

So the prompts are tuned and adjusted on a per-model basis. If you look at the number of attempts, each receives a specific prompt variation depending on the model. This honestly isn't as much of an issue these days because SOTA models natural language parsing (particularly the multimodal ones) has eliminated a lot of the byzantine syntax requirements of the SD/SDXL days. The template prompt seen in each comparison g…

Shouldn’t every model get the same prompt? Seems a bit weird, especially when you can’t see the prompts that were used.

Re: ChatGPT Images 2.0

#519

Post by vaulstein »

davebren wrote: post #52
simonw wrote: post #37

Earlier quoted context omitted.

I just got a much better version using this command instead, which uses the maximum image size according to https://github.com/openai/openai-cookbook/blob/main/examples... OPENAI_API_KEY="$(llm keys get openai)" \ uv run 'https://raw.githubusercontent.com/simonw/tools/refs/heads/main/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" \ --quali…

The faces...that's nice that it turned a kid's book into an abomination

It's interesting that the raccoon is well defined because it was a part of the request. But none of the other Fauna are.

Re: ChatGPT Images 2.0

#520

Post by jumploops »

Looks like analog clocks work well enough now, however it still struggles with left-handed people.

Overall, quite impressed with its continuity and agentic (i.e. research) features.

Post reply on HN