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

Generative AI Image Editing Showdown

genai-showdown.specr.net

Post reply on HN
Print view View original

61–70 of 83 posts

Re: Generative AI Image Editing Showdown

#61

Post by vunderba »

epiccoleman wrote: post #59
minimaxir wrote: post #4

Everyone is sleeping on Gemini 2.5 Flash Image / Nano Banana. As shown in the OP, it's substantially more powerful than most other models while at the same price-per-image, and due to its text encoder it can handle significantly larger and more nuanced prompts to get exactly what you want. I open-sourced a Python package for generating from it with examples ( https://github.com/minimaxir/gemimg ) and am currently wor…

half the time when i try to use nano banana, AI Studio fails, telling me it can't generate for some unspecified reason. these aren't cases where I'm trying to do something that skirts the edge of copyright, either (like "Ghiblifying" images, for example). that said, when it does work, it is super impressive.

It might be the safety moderation system. It's rather aggressive and when it does kick in (at least in the API), it often returns an empty response giving basically zero indication as to the root cause.

Re: Generative AI Image Editing Showdown

#63

Post by vunderba »

joomla199 wrote: post #3

Good effort, somewhat marred by poor prompting. Passing in “the tower in the image is leaning to the right,” for example, is a big mistake. That context is already in the image, and passing that as a prompt will only make the model apt to lean the tower in the result.

I should have been more clear. Those are NOT the direct prompts. They are the starter prompts. In fact that's why the attempt numbers change, we adapt the exact prompts depending on the model.

Re: Generative AI Image Editing Showdown

#64

Post by vunderba »

keyle wrote: post #10

This was fun. Some might critique the prompts and say this or that would have done better, but they were the kind of prompt your dad would type in not knowing how to push the right buttons.

OP here. You're the second person to say this. I cut my teeth on SD 1.5 - so I'm rather intimately familiar (for better or worse) with the level of prompt craft necessary depending on the model.

I feel like the FAQ section isn't displayed prominently enough:

How are the prompts written?

  In addition to giving models several attempts to generate an image, we also write several variations of the prompt to ensure that models don't get stuck on certain keywords or phrases depending on their training data. For example, while hippity hop is a relatively common name for the ball riding toy, it is also known as a space hopper. We try to use both terms in the prompts to ensure that models are not biased towards one or the other.

  Prompts for Hunyuan were attempted in both Chinese and English with and without Image Optimization.
Additionally when you see a prompt like "Turn on the lights" - the idea is to actually go beyond direct prompting commands - we're actually probing the capabilities of a truly multimodal LLM. It's a prompt that would spectacularly fail in more traditional models (such as SDXL).

Re: Generative AI Image Editing Showdown

#65

Post by joomla199 »

vunderba wrote: post #63
joomla199 wrote: post #3

Good effort, somewhat marred by poor prompting. Passing in “the tower in the image is leaning to the right,” for example, is a big mistake. That context is already in the image, and passing that as a prompt will only make the model apt to lean the tower in the result.

I should have been more clear. Those are NOT the direct prompts. They are the starter prompts. In fact that's why the attempt numbers change, we adapt the exact prompts depending on the model.

I understood that much, at least from the description you added on the Kontext result. I agree that you should provide more information here, though, especially around "we adapt the exact prompts depending on the model", since your strategy here could also reflect model strengths and weaknesses.

Re: Generative AI Image Editing Showdown

#66

Post by vunderba »

joomla199 wrote: post #65
vunderba wrote: post #63

Earlier quoted context omitted.

I should have been more clear. Those are NOT the direct prompts. They are the starter prompts. In fact that's why the attempt numbers change, we adapt the exact prompts depending on the model.

I understood that much, at least from the description you added on the Kontext result. I agree that you should provide more information here, though, especially around "we adapt the exact prompts depending on the model", since your strategy here could also reflect model strengths and weaknesses.

Good point! Perhaps I should add in the "final model-specific prompt", or place them in an errata section.

Re: Generative AI Image Editing Showdown

#67

Post by minimaxir »

franze wrote: post #58
minimaxir wrote: post #4

Everyone is sleeping on Gemini 2.5 Flash Image / Nano Banana. As shown in the OP, it's substantially more powerful than most other models while at the same price-per-image, and due to its text encoder it can handle significantly larger and more nuanced prompts to get exactly what you want. I open-sourced a Python package for generating from it with examples ( https://github.com/minimaxir/gemimg ) and am currently wor…

honest question: where is / how to do aspect ratio control for nano banana in aistudio?

It's on the right sidebar if Nano Banana is selected.

Re: Generative AI Image Editing Showdown

#68

Post by ec109685 »

troupo wrote: post #52
ec109685 wrote: post #39

Earlier quoted context omitted.

Google photos should just include the feature. It’s kinda buried in Gemini. Google is so weirdly non-integrated.

> It’s kinda buried in Gemini. > Google is so weirdly non-integrated. Where by try gemini non- integrated have you tried gemini you mean gemini is here they shove use gemini gemini into every single product they have?

It is terrible in all those services.

Re: Generative AI Image Editing Showdown

#70

Post by seany »

tpierce89 wrote: post #60
seany wrote: post #44

Is there anything like this comparison for nsfw images? I'm married to a boudoir photographer who sometimes wants to use ai tools for things, and they are all _awfull_ if there is nudity on photos. It's like some sort of neo puritanism has taken over.

I also do similar work and have run tests on many models. I have listed a few here with sample images using one prompt with a single run. I know it isn't a comprehensive review like OP, but it's something. My personal preference through experience is epicRealismXL. https://imgchest.com/p/xny8e23jpyb

Thanks for the tip. Need to see how well these work for inpainting
Post reply on HN