Live data from Hacker News

Gemini 2.5 Flash Image

developers.googleblog.com

331–340 of 504 posts

Re: Gemini 2.5 Flash Image

#331
post #48

Earlier quoted context omitted.

This presumes that you're okay with giving the real Elon your wallet but not a fake Elon, but why?

It was very convincing. We thought it was a YouTube stream of the Starship launch. It paused with 40 seconds remaining, and "Musk" came on offering to reward those who support innovation and technology (BTC, in this case). All info here: https://docs.google.com/document/d/1lRbApgKT4U95zN0AYsPQqsLR...

My problem with your statement isn't if its believable Elon came on stage or not, my problem is why would you trust Elon to pay you your money back, whether its the authentic or imposter Musk.

Re: Gemini 2.5 Flash Image

#332
post #48

Earlier quoted context omitted.

This presumes that you're okay with giving the real Elon your wallet but not a fake Elon, but why?

It was very convincing. We thought it was a YouTube stream of the Starship launch. It paused with 40 seconds remaining, and "Musk" came on offering to reward those who support innovation and technology (BTC, in this case). All info here: https://docs.google.com/document/d/1lRbApgKT4U95zN0AYsPQqsLR...

Kind of missing their point there. Giving Elon Musk $15k in crypto based on some vague too-good-to-be-true "trust me bro" pitch is embarrassing even if the video turned out to be real.

Re: Gemini 2.5 Flash Image

#333
post #60

This is the gpt 4 moment for image editing models. Nano banana aka gemini 2.5 flash is insanely good. It made a 171 elo point jump in lmarena! Just search nano banana on Twitter to see the crazy results. An example. https://x.com/D_studioproject/status/1958019251178267111

> This is the gpt 4 moment for image editing models. No it's not. We've had rich editing capabilities since gpt-image-1, this is just faster and looks better than the (endearingly? called) "piss filter". Flux Kontext, SeedEdit, and Qwen Edit are all also image editing models that are robustly capable. Qwen Edit especially. Flux Kontext and Qwen are also possible to fine tune and run locally. Qwen (and its video gen s…

I'm sorry I absolutely don't agree. This model is on a whole other level.

It's not even close. https://twitter.com/fareszr/status/1960436757822103721

Re: Gemini 2.5 Flash Image

#334

Earlier quoted context omitted.

User has edited their bio now :)

I didn't edit my bio. My projects are not satire. I'm just less ashamed than most, so I work on more "exciting" projects. I've worked extensively with generative AI, including video, myself. It was just that convincing to me in the moment. My regret knows no bounds. Luckily I earn enough this doesn't devastate me, but I really could have done some good with that money.

Yikes. In that case, please accept my apology. Your bio disappeared for a while off your page, but it's back as it was now.

Re: Gemini 2.5 Flash Image

#335

Super cheap generation but expensive image upload, do I read that right? https://openrouter.ai/google/gemini-2.5-flash-image-preview

Not sure. If the Flash image output is $30/M [1] then that's pretty similar to gpt-image-1 costs. So a faster and better model perhaps but not really cheaper? [1] https://developers.googleblog.com/en/introducing-gemini-2-5-...

Since I can't edit, it seems like Flash image is about 23% (4 cents vs 17 cents) of the cost of Openai gpt-image-1, if you're putting an image and prompt in and getting out, say, a 1024x1024 generated image. With the quicker production time that makes it interesting. Expecting Openai to respond at least in terms of pricing, e.g. a flat rate output cap price or something to be comparable.

Re: Gemini 2.5 Flash Image

#336
post #60

This is the gpt 4 moment for image editing models. Nano banana aka gemini 2.5 flash is insanely good. It made a 171 elo point jump in lmarena! Just search nano banana on Twitter to see the crazy results. An example. https://x.com/D_studioproject/status/1958019251178267111

Regardless, it seems Google is on the frontier of every type of model and robotics (cars). It’s nutty how we forget what a intellectual juggernaut they are.

Re: Gemini 2.5 Flash Image

#337

Earlier quoted context omitted.

If you compare to the amount of effort required in Photoshop to achieve the same results, still a vast improvement

Vibe coding might not be real, but vibe graphics design certainly is. https://imgur.com/a/internet-DWzJ26B Anyone can make images and video now.

What tools did you use to make those videos from the PG image?

Re: Gemini 2.5 Flash Image

#338
A bit mixed opinions - I tried colorizing manga pages with it, and the results were perfect.

Interestingly, it can change pages with tons of text on them without any problem, but cannot seem to do translation, if I ask it to translate a French comic page, the text ends up garbled (even though it can perfectly read and translate the text by itself).

I tried with another page, and it copypasted the same character (in different poses!) all over the panels. Absolutely uncanny!

However when I asked to remake a Western comic book in a manga style (provided a very similar manga page to the comic one), it totally failed.

Also about 50% of the time, it just tells me it'll generate the image but doesn't actually do it - not sure what's going on but a retry fixes it, but it's annoying.

Re: Gemini 2.5 Flash Image

#340
post #60

This is the gpt 4 moment for image editing models. Nano banana aka gemini 2.5 flash is insanely good. It made a 171 elo point jump in lmarena! Just search nano banana on Twitter to see the crazy results. An example. https://x.com/D_studioproject/status/1958019251178267111

cope
Post reply on HN