How to convert your photos into studio ghibli anime style using AI

#ai #ghibli

2025-04-01 Edit:
ChatGPT now allows free users to access its image generation feature, but with rate limits(3 images per day maybe). If you want a more intuitive UI and faster processing options, you can still try this website: https://www.ai-anime-image.com/

2025-03-27 Edit:
It seems that OpenAI has not delivered the new ability to free users. So if you have a plus, pro or team subscription plan, you are good to go. But if your are a free plan user. You can not use this in ChatGPT. I have found a website providing the same functionality, try it out: https://www.ai-anime-image.com/

Open AI release their image generation feature for gpt-4o model. The image generation is really awesome. And the model can convert photos into ghibli anime style very perfect. This has tract a lot of attention on x.com.

Now let's me tell you how to convert photos into ghibli anime style.

According to OpenAI's announcement, the new model is rolling out, so you might not get access to the latest model. If openAI has not make the latest model avaiblable to you, you can only use previous DALL·E model.

You can try upload a photo and use this prompt to get the ghibli anime style photo.

convert this photo to studio ghibli style anime

demo:

And there are many other prompt you can try:

Make this into Ghibli style anime

make it japanese comic style with high resolution and detailed artwork.

Please convert this image into an anime-style illustration inspired by the visual aesthetics of Studio Ghibli. Maintain the original composition andkey elements, but reimagine them with Ghibli-style features - soft, painterly textures, warm and natural color palettes, detailedbackgrounds, and expressive character designs. Aim for a whimsical andnostalgic atmosphere, capturing the magical realism often seen in Ghiblifilms like My Neighbor Totoro, Spirited Away, or Howl's Moving Castle.

Interested in the latest AI tools? Discover the most advanced ones in our AI tools index. 🚀

References

Introducing 4o Image Generation | OpenAI

Top comments (29)

Nelson Figueroa • Mar 26

I wonder how Studio Ghibli feels about their creative works being stolen to train AI.

Andy • Mar 27

Yeah, yesterday I just thought it was fun, but today I’ve already seen a lot of news reports discussing copyright issues. Hayao Miyazaki has previously expressed his dislike for using AI to create animation, so I wonder how Studio Ghibli will respond this time.

Nelson Figueroa • Mar 27

Yep! I think this is cool from a tech point of view but ethically it's not so cool.

Ben Sinclair • Mar 30

Given (a) what they've said in the past and (b) how most artists feel about this sort of thing, I suspect they are not happy.

Daan • Mar 27

Crazy how they got stolen from by losing nothing and own things they didn't make.. When will people drop this terrible reasoning?

Nelson Figueroa • Mar 30

You are misinformed if you think Studio Ghibli loses nothing.
Studio Ghibli doesn't have to lose anything for this to be unethical.

Also, you really created an account just to leave this comment? lol

Daan • Mar 31 • Edited

It's like Nintendo complaining on losing business from Palworld vs Pokémon. Some might argue otherwise but "copyright on styles" was a bad idea to begin with and is especially unmaintainable nowadays. If Studio Ghibli and Nintendo are inherently better than AI art and Palword, they have nothing to fear nor lose, and do not require cease and desists to come out ahead.

If the claim of unethical is based on it being theft, then they absolutely require to lose something or the entire argument crumbles.

Also, yes, I never heard of this website before reading this post, so thanks for the lukewarm welcome <3

Dushmanta • Mar 27

Is it for pro users only? it didn't work the response said it didn't have the capabilities

Andy • Mar 27

OpenAI might be rolling out this new feature in batches, so Plus, Pro, and Team users could get access earlier, while free users may have to wait until OpenAI gradually expands availability. If the new feature hasn’t been enabled for your account yet, ChatGPT will default to using the DALL·E model for image generation, which could result in refusals or lower-quality outputs compared to the new model.