Nano Banana Pro Text-to-Image Multi
Generate two 2K images from one prompt with Nano Banana Pro Multi on Conut.ai. Four aspect ratios, PNG or JPEG, 20 credits for one image and 40 for two.
Nano Banana Pro Multi generates two 2K images from a single prompt in one run. It is built for comparison: two takes on the same brief, side by side, before you commit to a direction. One image costs 20 credits and two cost 40, so a pair costs the same as one standard Nano Banana Pro image. The trade is control — Multi runs at 2K with four aspect ratios, against ten on the standard page — but for early decisions that is usually enough.
- Images per run: 1 or 2
- Resolution: 2K
- Price: 20–40 credits
- Aspect ratios: 4
How to use Nano Banana Pro Text to Image Multi
Write one prompt, request two images, then pick the stronger variation instead of rerolling the whole idea.
- Write a prompt that can vary Leave some room in the phrasing: a pose, a camera angle or a background detail can differ without breaking the brief. Over-specifying every element makes both images nearly identical.
- Choose the aspect ratio Only 3:2, 2:3, 3:4 and 4:3 are available here. Pick the final frame before generating, since the crop is fixed per run.
- Set one or two images Two images cost 40 credits and are usually the better deal when you are choosing a direction. One image is enough when the prompt is already settled.
- Compare at the same zoom Judge both results at full size on the same screen. Look for the details that matter to the brief rather than picking the brighter image.
- Carry the winner forward Use the chosen image as a reference on the Edit page for fixes, or rerun the prompt on Ultra when the file needs more resolution.
Settings reference
Multi returns one or two 2K images from a single prompt, which makes it useful for choosing between variations without writing a second prompt.
| Setting | Value | Notes |
|---|---|---|
| Images per run | 1 or 2 | Default 2. Both images come from the same prompt and seed-independent sampling. |
| Price | 20 credits each | 20 for one image, 40 for two. |
| Resolution | 2K | Fixed on this page. Use Ultra for 4K or 8K. |
| Aspect ratio | 4 options | 3:2, 2:3, 3:4 and 4:3; default 3:4. |
| Output format | JPEG / PNG | Choose PNG when either result may be edited further. |
| Reference images | None | Text-only. Use Edit Multi to generate variations from an uploaded image. |
What Multi is for
Choosing between two directions is easier when the two images share everything except the variable you care about. Multi gives you that comparison in one run: same prompt, same settings, two outputs that differ only in the model's own variation.
It is also a useful client-facing tool. Instead of sending one option and a revision, send two takes and let the choice happen before the high-resolution pass.
- Compare two takes on one brief
- Present options to a client in one round
- Test a prompt before scaling to 4K or 8K
- Fill a two-image layout with one generation
- Explore composition without rewriting the prompt
Settings and credits
Multi is deliberately narrow: 2K only, four aspect ratios, no reference upload. The image count is the only variable that changes the price.
| Setting | Options | Notes |
|---|---|---|
| Images per run | 1 or 2 | Default 2 |
| Resolution | 2K | Fixed |
| Aspect ratio | 3:2, 2:3, 3:4, 4:3 | Default 3:4 |
| Output format | PNG / JPEG | PNG for graphics and text |
| Credits | 20 for 1 / 40 for 2 | Per run |
| Reference images | None | Use Nano Banana Pro Edit Multi to edit |
How the two images differ
Both images come from the same prompt, so the difference is the model's internal variation, not two separate briefs. That is what makes the pair useful: you see which parts of the composition are stable and which are not.
If the two outputs look too similar, change one word or one setting and run again. If they look too different, your prompt is under-specified — add the invariants you actually need, such as framing, light or the subject's position.
- Write one strong prompt, not two
- Lock the aspect ratio before comparing
- Read the pair for stable versus unstable details
- Pick one and refine it on the standard or Ultra page
- Use 1 image when you only need a single take
Multi vs standard vs Ultra
Standard gives one image at 2K for 40 credits across ten ratios. Multi gives one at 20 or two at 40, but only four ratios. Ultra gives one at 4K or 8K for 40 or 50. Multi is the cheapest way to see two results; Ultra is the right page once a direction is chosen and resolution matters.
| Page | Output | Ratios | Credits |
|---|---|---|---|
| Nano Banana Pro (standard) | 1 × 2K | 10 | 40 |
| Nano Banana Pro Multi (this page) | 1 or 2 × 2K | 4 | 20 / 40 |
| Nano Banana Pro Ultra | 1 × 4K or 8K | 10 | 40 / 50 |
| Nano Banana 2 | 1 × 1K / 2K / 4K | 10 | 30 / 40 / 50 |
Limits to plan around
- Only four aspect ratios: 3:2, 2:3, 3:4 and 4:3
- 2K output only; no 4K or 8K on this page
- No reference images
- Maximum of two images per run
- Each run costs 20 or 40 credits depending on image count
Prompt examples for Nano Banana Pro Text-to-Image Multi
- Product direction: A glass perfume bottle on a pale sand plinth, hard afternoon sunlight casting a crisp shadow, a single dried grass stem beside it, warm neutral background, minimal luxury product photography, 1:1.
- Portrait options: A quiet editorial portrait of a woman in a grey knit coat, standing against a weathered concrete wall, soft overcast light, calm expression, 85 mm lens, muted palette, 4:5.
- Poster concept: A minimalist film poster for a desert road movie, a single empty highway vanishing into heat haze, burnt orange sky, small white title block at the bottom, grainy 1970s print texture, 2:3.
- Interior mood: A serene reading corner with a linen armchair, a low oak side table and a stack of books, warm lamp light against a deep blue evening window, soft shadows, 3:4.
- Character design: A full-body character design of a desert nomad in layered linen wraps, brass jewellery and worn leather boots, neutral stance, clean studio background, concept art turnaround lighting, 3:4.
- Food styling: A rustic bowl of ramen on a dark wooden table, steam rising, soft-boiled egg and scallions, overhead shot, moody side light, deep shadows, editorial food photography, 3:2.
Pro tips for Multi
Two images are most useful when you compare a real decision, not two near-duplicates.
- Ask for variation in the prompt Mention that the subject may be shown from a different angle or with a different expression. The model then uses the second image to explore.
- Do not mix two ideas One prompt should still describe one scene. Two unrelated concepts produce two weak images instead of two options.
- Use two images for A/B briefs When a client needs options, the two-image run is cheaper than two separate generations and keeps the same visual language.
- Judge identity first If the image contains a person or product, compare faces and shapes before style. A prettier image that changes the subject is not the winner.
- Finish on Ultra or Edit Multi caps at 2K and four ratios. Move the selected image to Ultra for resolution or Edit for targeted changes.
Frequently asked questions
What is Nano Banana Pro Multi?
It is the multi-image page for Google's Nano Banana Pro on Conut.ai. One prompt produces one or two 2K images in a single run.
How much does a run cost?
20 credits for one image and 40 credits for two.
What resolution and ratios are available?
2K only, in four aspect ratios: 3:2, 2:3, 3:4 and 4:3.
Are the two images different?
Yes, they are two samples from the same prompt. The differences show which parts of the composition are stable.
Can I upload reference images?
Not on this page. Use Nano Banana Pro Edit Multi for reference-based variation.
Which formats are available?
PNG and JPEG.
Why is Multi cheaper than the standard page?
Multi is priced per image at 20 credits, with a narrower feature set: 2K only and four aspect ratios instead of ten.
Compare with other image models
- Nano Banana Pro Text to Image
- Nano Banana Pro Ultra
- Nano Banana Pro Edit Multi
- Nano Banana 2 Text to Image
- GPT Image 2 Text to Image
- Browse all image models
Specifications
- Provider: Google
- Model type: Text to image
- Credit pricing: 20 credits per image
- Capabilities: starter, multi-shot
About Conut.ai
Conut.ai is an AI image, video and audio generation platform with 50+ creative tools. It supports text-to-image, image-to-video, face swap, motion control, relighting, inpainting and video effects through models including Nano Banana Pro, Seedream, GPT Image, Sora 2, Kling, WAN and Seedance. Start free — no credit card required.