The images generated by GPT before were really uncomfortable, mainly because the Chinese characters were always wrong! Correcting them was no use!
Then I looked it up and found out that #ChatGPT Images 2.0 has been released!
This time, besides supporting Chinese, there's also a major editable feature!
I just used it to give a command, asking it to generate what it thinks are the most beautiful: Chinese, Japanese, and Korean beauties, and removed the AI vibe~
Take a look at how the results of ChatGPT Images 2.0 turned out this time 👇

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

12 Likes

Reward
12
15
8
Share

Comment

Add a comment

QuietValidator

· 04-29 01:33

What is the current level of support for Chinese typesetting? Are vertical writing, traditional characters, and punctuation spacing still stable?

View OriginalReply0

FlamingoFrontView

· 04-27 05:37

This AI effect is quite impressive, but it also depends on whether the lighting, shadows, and skin textures are overly smoothed.

View OriginalReply0

OrderbookOtter

· 04-26 20:53

It really hits harder than before, and Chinese characters finally aren't acting up anymore.

View OriginalReply0

ThereAreCatsInTheContract.

· 04-26 20:25

I want to see a comparison chart between Gemini and Images 2.0 under the same prompt set; the difference will be more intuitive.

View OriginalReply0

Stop-LossLineForTheEveningGlow

· 04-26 19:10

The editing feature is so awesome; previously, after generating, I could only start over, but now I can tweak details and save a lot of time.

View OriginalReply0

Stop-LossIsLikeAConfession

· 04-26 18:59

I'm a bit concerned about copyright style issues; try not to imitate specific photographers or celebrity faces, or it might cause problems.

View OriginalReply0

TheAccountHasBeenBanned

· 04-26 18:55

Gemini's visuals have always felt "clean but flat" to me, but GPT now has more texture and depth, especially in the lighting and shadows.

View OriginalReply0

GateUser-0c86a4c9

· 04-26 18:51

Does this set you generated have the old issues like fingers/ear jewelry/hair strands? When zooming in on the details, it's easiest to mess up.

View OriginalReply0

PuddingMarketMaker

· 04-26 18:42

Looking forward to your follow-up test: Will re-editing the same image (changing background, modifying text, changing expressions) result in it becoming more blurry the more you edit?

View OriginalReply0

OnChainMonk

· 04-26 18:41

I am more concerned about consistency: can the same person's face remain recognizable across multiple images, different poses, and changing scenes?

View OriginalReply0

Trending Topics
View More
#
SummerCreationCamp
1.51M Popularity
#
EventContractsLaunch
137.05K Popularity
#
BrentReturnsTo100
1.56M Popularity
#
IntelQ2RevenueSurges25%
411.69K Popularity
#
UStoImpose10To12.5PercentTariffsOn60Economies
11.81M Popularity

Pinned

Sitemap

I've always used Gemini mainly for writing articles and creating images, until a few days ago when I accidentally used ChatGPT to make an image and realized it was much better than before!

Trending Topics

SummerCreationCamp

EventContractsLaunch

BrentReturnsTo100

IntelQ2RevenueSurges25%

UStoImpose10To12.5PercentTariffsOn60Economies

Pinned