Gemini (Nano Banana) and Flux Kontext Pro models now in Photoshop MAX update 27.0

[Srishti Bali]

Big news for Photoshop creators! 

We’re thrilled to announce that Gemini (Nano Banana) and Flux Kontext Pro models are now integrated into the Photoshop Desktop Generative Fill workflow! These models unlock advanced capabilities to help with your creative vision, including:

• Style transfer
• Text editing
• Object modification
• Quick or controlled edits
• Complex transformations

All within a non-destructive, creative control environment.

Try it out now Download the latest version and dive into the new Adobe Photoshop features.

Update Photoshop 27.0

To update Photoshop 27.0, click "Update" in the Creative Cloud desktop app next to Photoshop. More detailed instructions for updating.

Here's how to use these Models in Photoshop's Generative Fill:

• Open a file → Create a selection (try Select All or Selection Brush)
• Click the Generative Fill button
• Tap the Model Picker icon
• Choose an AI model. 
• Enter a detailed prompt → Generate!
• Click here for more details.

Pro-Tips for using Generative Fill:

• Use larger, loose selections with the Selection Brush.
• Avoid hard edges to prevent shifting or visible seams.
• Add clear prompts to guide the model’s output.
• Use Flux Kontext for seamless, context-aware backgrounds. It is conversational, helping maintain character consistency and precise edits.
• Use Nano Banana for stylistic fills. It is optimized for preserving details, making it ideal for creative transformations.

Important Note About Generation Credits:

• Generative Fill and Generative Expand using Adobe Firefly models are standard features and will consume 1 credit per generation.
• Generative Fill and Generative Expand, powered by Firefly Fill & Expand (beta) model in Photoshop (beta) on desktop, deduct 1 credit per generation.  
• Generative Fill, powered by powered by partner AI models, are premium features:
  • Gemini 3 (with Nano Banana Pro) = 40 credits
  • Gemini 2.5 (Nano Banana) = 10 credits
  • FLUX.2 pro = 20 credits
  • FLUX.1 Kontext [pro] = 10 credits

Learn more about premium features and generative credits in the generative credits FAQ.

Are you ready to get started? Update Photoshop to the latest version and explore these LLM models and other new exciting features. More details can be found here.

We’d love to hear from you!

Try out these new models in Photoshop 27.0 and let us know your thoughts. Feel free to reach out if you have any questions or need more tips!

If this was helpful, please click the Upvote icon in the top left corner of this post. Your feedback helps us improve the experience. Thank you!

We’d love to see what you’ve created and hear about your experience or feedback! Click 'Reply' below.

Click 'Follow' below to bookmark this post for easy reference.

For other questions, feel free to start a new post in the community.

[onazole]

I'm not a premium user and I've been able to use generative fill with Nano Banana and Flux. I'd like to know if the maximum of 100 uses per month that was in the beta is also available now in version 27.

Thanks.

[Sarah_Jane0168]

I’m also curious about this. I’ve been using generative fill without a premium plan, and it’s been working great so far. Just wondering if the usage limit from the beta still applies in version 27 or if it’s changed.

[CREAM Motion]

I was very excited to have Nano Banana in Photoshop, but it doesn't work for me!
There is always a discrepancy between the original image and the generated image...
I used up a hundred credits for nothing.

 

  

[Glenn 8675309]

That's the same result I just got- it looked like a cutout from a paper doll set. 

[fariya_yousaf]

This is a very powerful update. It's exciting to see advanced models like Gemini 2.5 and Flux directly integrated.

The 10-credit cost per generation for these new models is a steep jump, which will make experimentation more considered compared to the 1-credit Firefly.

That said, the pro-tip about using Flux for 'conversational' character consistency sounds like it could be a massive time-saver for complex projects.

Looking forward to testing if the performance justifies the cost. Thanks for the detailed breakdown!

[ali_9656]

Amazing update! The integration of Gemini 2.5 (Nano Banana) and Flux Kontext Pro really enhances the Generative Fill workflow. I’ve tested both models and the results are noticeably smoother and more accurate. Great job, Adobe team — love how fast these AI improvements are evolving!

[h3px1]

As a Gemini Pro user, I naturally assumed I could use Nano Banana as part of my Adobe and Gemini subscriptions without having to pay Adobe again. Now I'm supposed to pay a third time. I don't think that will work for me.

[onazole]

And the Harmonize option costs five credits.

[anamtaq]

will look into this

[andy_1823]

This update sounds amazing! I'm excited to try the Gemini 2.5 and Flux Kontext Pro models in Photoshop. The new style transfer features could really boost my creative projects. Thanks for the helpful news!

[Glenn 8675309]

I tired the flux Kontext last night- I either got one 2 tries- then they want to stick you with a subscription for that feature.

[Roni_Erric6360]

Excited to try the Gemini 2.5 and Flux Kontext Pro models in the latest Photoshop MAX update. Looks like some powerful new tools.

[Glenn 8675309]

Only getting one or two uses before you to subscribe to yet another thing from Adobe makes it virtusally useless for the non -business user of the sfotware.

[Cap Horizon]

Hello,
I've used up all my Nano Banana and Flux.1 Kontext credits. I'm only being offered Firely credits, even though I have 1000/1000! How can I get more Nano Banana and Flux.1 Kontext credits?

[Noel Orridge]

Hi @Cap Horizon! Thank you for reaching out! Could you please send me your email address privately, so I can look into this further? Thanks! 

[Cap Horizon]

Hi
There is my email adresse below.
Sincerely yours.
Rémy Cantin

[PII removed by moderator]

[Calla Luna]

I am on an enterprise account and our admin/IT hasn't approved the Parner Model functionality within Photoshop yet. As a workaround, can I use an adobe stock photo in nano banana to create a video clip and use it commercially?

[ambrosial_curiosity7010]

• f the original Adobe Stock image is licensed for commercial use → your final output is fine for commercial use.

• If the original is only for personal/editorial use → then the final output also can’t be used commercially.

[eimmaarose]

Thanks for the update!
The new integration of Gemini 2.5 (Nano Banana) and Flux Kontext Pro into Photoshop 27.0 looks great so far.

I’ve tested a few features already, and here’s what I noticed:

Flux Kontext Pro handles complex edits with much better structure and realism—especially for object manipulation.

Gemini 2.5 feels noticeably faster for quick creative adjustments and texture-style edits.

The Model Picker inside Generative Fill is a big improvement. Switching models for different tasks makes the workflow smoother.

One question:
Are there any recommended prompt styles or optimization tips from Adobe for getting the best results from each model? Especially for text editing and style transfer?

Overall, loving the direction—thanks for continuing to improve the GenAI workflow in Photoshop!

[Cap Horizon]

I have 1000 Firely credits per month. Does this mean they cannot be used to renew Nano Banana or Fux.1 Kontext credits?

[chloe_4778]

Thanks for sharing this guide! I’m new here and excited to try out Gemini 2.5 (Nano Banana) and Flux Kontext Pro in Photoshop Max.
The pro tips and credit info are very helpful. I’ll follow them while experimenting and look forward to learning from others’ results and sharing my own soon.

[Briefcase_Joe2394]

Huge news! Generative Fill was already great, but adding models optimized for consistency (Flux) and detail (Nano Banana) is next level. Congrats to the team on getting these integrated so quickly. Downloading the update now to test them out!

[Hanory]

Thanks for the update! It’s exciting to see Flux Kontext Pro and Gemini 2.5 integrated into Generative Fill, especially with different strengths for context-aware edits vs. stylistic transformations. I do have one question though—do these models use the actual canvas context the same way Firefly does, or do they rely mostly on the prompt and selection area? Just trying to understand how precise the control is.