Gemini 2.5 Flash Image (Nano Banana) best practices 🍌🍌🍌 - Be hyper-specific: The more detail you provide, the more control you have. Instead of "fantasy armor," describe it "ornate elven plate armor, etched with silver leaf patterns, with a high collar and pauldrons shaped like falcon wings." - Fix character consistency drifts: If you notice a character's features begin to drift after many iterative edits, you can restart a new conversation with a detailed description to retain consistency. - Provide context and intent: Explain the *purpose* of the image. For example, "Create a logo for a high-end, minimalist skincare brand" will yield better results than just "Create a logo." - Iterate and refine: Don't expect a perfect image on the first try. Use the conversational nature of the model to make small changes. Follow up with prompts like, "That's great, but can you make the lighting a bit warmer?" or "Keep everything the same, but change the character's expression to be more serious." - Use "semantic negative prompts": Instead of saying "no cars," describe the desired scene positively, "an empty, deserted street with no signs of traffic." - Aspect ratios: When editing, Gemini 2.5 Flash Image generally preserves the input image's aspect ratio. If you upload multiple images with different aspect ratios, the model will adopt the aspect ratio of the *last* image provided. - Control the camera: Use photographic and cinematic language to control the composition. Terms like `wide-angle shot`, `macro shot`, `low-angle perspective`, `85mm portrait lens`, and `Dutch angle` give you precise control over the final image.

Aug 30, 2025 · 3:12 PM UTC

64
244
22
2,016
Replying to @_philschmid
Providing the context and intent has been really helpful in getting perfect images.
4
Replying to @_philschmid
Again lazy prompt will not take you anywhere
1
3
Replying to @_philschmid
If you’re looking for a diverse range of shot types, this may help.
C I N E M A T I C . C A M E R A . P R O M P T S // T E X T - T O - V I D E O . T H R E A D @runwayml @Hailuo_AI @LumaLabsAI @pika_labs @Kling_ai @Viduforhuman
3
Replying to @_philschmid
We got Nano Banana agents in Coral too 😎
2
Replying to @_philschmid
Control emerges from specificity. But too much detail can stifle imagination. Find the balance.
Replying to @_philschmid
Super Helpful, Philip!
Replying to @_philschmid
the aspect ratio thing was driving me crazy ty! is there any other guidance for enforcing aspect ratio? just asking in the prompt doesn’t seem to do it but i was thinking a hack could just be to pass a blank image in the ratio i want as the last image and describe my edits with the other inputs
2
9
Replying to @_philschmid
why are we giving Ai engineers a pass on the pisspoor natural language inferences and requiring users to learn these best practices.. Ai is meant to be a tool for humans to understand human language and expectations. We mold it to our whims not change our ways for Ai.
4
1
5
Replying to @_philschmid
9wEkKb76AUZLfWTDGuKxVAw54tX6RneTkKsA7XNwpump
3
Replying to @_philschmid
It would be really valuable to know if it's possible or not to end up in a "Shadow Ban"-Mode. Feels from my usage and reading of reports like it.
Replying to @levelsio
For a joke purpose I've tried to get via aistudio.google.com and Gemini to get Nano Banana to combine a picture of myself and the QAnon Shaman so I wear his getup. I couldn't get it to do it. It refused a bunch of times and when it did, it felt like it just didn't even try.
2
Replying to @_philschmid
9wEkKb76AUZLfWTDGuKxVAw54tX6RneTkKsA7XNwpump
2
Replying to @_philschmid
Are these applicable to Veo 3 as well?
1
Replying to @_philschmid
Thanks for sharing mate! 🙌
1
Replying to @_philschmid
agreed, follows a similar structure to prompting with Veo3 i think many are used to MJ doing the heavy lifting on their image prompts best to be detailed with NB put words to what you see in your mind
1
Replying to @_philschmid
It's the lack of specificity that kills us. We expect the AI to understand everything that we are thinking.
1
Replying to @_philschmid
Nice! Long live the Banana
1
Replying to @_philschmid
At the end of the day , it all boils down to domain knowledge. The better you have it, the better you know what to prompt and which key words to use. I think this will not make a lot of jobs disappear , cause most of them require domain knowledge 😅
1
Replying to @_philschmid
Any way to specify the output format ? Dimension and file type ?
1
Replying to @_philschmid
9wEkKb76AUZLfWTDGuKxVAw54tX6RneTkKsA7XNwpump This
1
Replying to @_philschmid
Literally "The Basics of Prompt Engineering" Applicable not only to the Nano Banana thing
1
Replying to @_philschmid
9wEkKb76AUZLfWTDGuKxVAw54tX6RneTkKsA7XNwpump
1
Replying to @_philschmid
BANANAROOM ca: 9wEkKb76AUZLfWTDGuKxVAw54tX6RneTkKsA7XNwpump
1
Replying to @_philschmid
Checks out with my experience!
Replying to @_philschmid
that's very helpful list of tips! 🙏
Replying to @_philschmid
Period correct camera
Tough times ahead for Image Editors like Photoshop if you can be creative and productive just with prompting. New Gemini 2.5 Flash is impressive!
Replying to @_philschmid
These are solid fundamentals that apply broadly across AI image generation models. The emphasis on specificity and iterative refinement is particularly valuable - many users underestimate how much precision in language translates to better visual outputs. The point about semantic negative prompts is especially insightful. Framing constraints positively rather than as exclusions tends to produce more coherent results across most diffusion models, not just Gemini. One addition I'd suggest: establishing a consistent 'visual vocabulary' early in a session can help maintain coherence across multiple generations. When you find terminology that produces the aesthetic you want, documenting and reusing those specific phrases creates more predictable results. The camera control techniques you mention are spot-on for anyone coming from a photography background - these models have clearly been trained on vast amounts of professionally shot and described imagery.
Replying to @_philschmid
Something I learned two years ago before any of this was a trend
Replying to @_philschmid
How can we reduce text errors when prompting nano banana, any tips?
Replying to @_philschmid
You can store specific details in a condensed format for images/videos using JSON prompting
Replying to @_philschmid
within minutes of nano banana going viral google frantically stumbled out the front door pushing the everyone aside, wild eyed, and stuttering, “guys!? the akchual name?! ..is Gemini Flash 2.5 Image!”
Replying to @_philschmid
The camera control tips are pure gold. It is like finally getting a manual for a spaceship.
Replying to @_philschmid
For best results, focus on intricate details like "vibrant yellow hue with subtle speckles, elongated shape, and a glossy finish."
Replying to @_philschmid
At this point, all of the technical terms that we have to specify in natural language should have been converted into buttons and UI/UX we can easily select and click. Interestingly, we are back to the time why invented Adobe Photoshop 😀