Nano Banana 2.1 Text to Image API

google/nano-banana-2.1/text-to-image
1K / 2K / 4K · 4 / 6 / 10 credits

Nano Banana 2.1 Text to Image turns text prompts into 1K–4K images with text layout rendering and panoramic compositions. It preserves perspective and lighting while eliminating tiling artifacts in panoramic aspect ratios.

Get API Key
Input
1230 / 20,000
OutputReady
6 credits / image · $0.030

Continue with

Examples

output.jpg

Create a sweeping 21:9 cinematic fantasy environment concept painting titled only in the viewer's imagination: Echo Survey of the Lava Fjord. An adult female geological surveyor in a weathered ochre expedition cloak leads one shaggy six-horned pack beast along a curved black basalt shore, both readable silhouettes in the lower left foreground. Their surveying equipment is practical and unbranded. Irregular glassy tidal pools and sharp wet volcanic stones fill the close foreground; a single continuous fjord sweeps through the middle distance toward a vast dormant volcano beneath layered low clouds on the far right. A restrained ember-red glow escapes one distant lava fissure while pale silver dusk lights the water and warm reflections touch the travelers. Build the entire panorama as one spatially coherent landscape with varied geological formations, natural transitions, consistent perspective, atmospheric distance, and a clear focal journey. No mirrored scenery, repeated stitched sections, duplicated travelers, or tiled rock patterns. Luminous matte-painting brushwork, epic scale, restrained color, a quiet human story within immense geology. No text, diagrams, labels, advertising, brands, logos, or signatures.

output.jpg

Paint a vertical 3:4 cinematic fantasy narrative keyframe inside the backstage dressing room of an underground theater carved into ancient dark stone. One elderly male tailor with small round spectacles carefully places the final stitch in a deep plum feather shawl worn by one adult female actress, moments before she enters the stage. Frame both expressive faces and the tailor's working hands clearly; show a single needle, a fine loose thread, layered iridescent feathers, worn velvet, and the actress gently holding one edge of the shawl. In the background, a narrow mirror shows a physically consistent partial reflection of the actress, a few hanging wooden puppet props remain distinct, and a gap in a heavy curtain reveals warm copper stage sidelighting against cool blue light entering through a high stone opening. Her anxious anticipation and his quiet concentration tell the story without words. Finished painterly concept art with refined gouache-like textures, precise material separation, believable anatomy, cinematic value grouping, rich shadows, and controlled detail. Exactly two main people; no extra limbs, duplicated tools, impossible mirror faces, lettering, advertising, brands, logos, or signatures.

Nano Banana 2.1 Text to Image

Nano Banana 2.1 Text to Image is an efficient image generation model developed by Google on Gemini 3.6 Flash. It balances fast inference with image quality, offering 1K, 2K, and 4K output with strong multilingual text layout rendering and panoramic artifact removal. The model delivers professional image generation at an accessible cost for poster design, infographics, and frequent commercial visual production.

Why Choose Nano Banana 2.1 Text to Image?

  • Native 1K, 2K, and 4K resolution optionsChoose native 1K (default), 2K, or 4K output to meet different delivery requirements, from everyday social media assets to professional print catalogs.

  • Advanced text rendering and complex layoutsEnhanced multilingual text generation and infographic layout capabilities produce legible, correctly spelled lettering for promotional posters, packaging labels, and signage.

  • Eliminates tiling artifacts in extreme panoramic ratiosFixes repeating patterns and stitching defects in ultrawide and extremely narrow panoramic formats such as 1:4, 4:1, 1:8, and 8:1, producing smooth, coherent compositions in formats including 21:9.

  • Flash inference speed and accessible pricingBuilt on the efficient inference architecture of Gemini 3.6 Flash, it substantially reduces generation costs compared with the previous generation. A 1K image costs just 4 credits ($0.020), supporting frequent creation and deployment at scale.

  • Multimodal thinking for stronger spatial depthBuilt-in multimodal Thinking capabilities provide strong instruction adherence and visual realism when handling long prompts with multiple objects, complex lighting, and spatial depth.

Parameters

ParameterRequirementDescription
promptRequired

Non-blank text, up to 20,000 characters.

sizeOptional

Output aspect ratio.

Default1:12:33:23:44:34:55:49:1616:921:9
resolutionOptional

Output resolution, which determines the generation price.

Default1K2K4K

How to Use

  1. Write a structured promptDescribe the subject, artistic style, lighting, and spatial composition in detail. Use double quotation marks to clearly identify any text that should appear in the image.

  2. Choose the aspect ratio and resolutionSelect one of 10 aspect ratios, such as 1:1 or 16:9, to suit the publishing format, then choose 1K, 2K, or 4K output resolution.

  3. Submit an asynchronous generation taskSend an API request with your parameters to receive a unique task_id. The backend cluster begins efficient parallel inference immediately.

  4. Poll the task statusUse task_id to query the generation status endpoint. Retrieve the returned image file list once the status changes to finished.

  5. Retrieve and download the finished imageRead the image URL from the response and save the finished output. An embedded SynthID digital identifier supports provenance tracking.

Pricing

One output image per request, billed by resolution. 1 credit = $0.005.

UsageRateDetails
1K4 credits / image$0.020 / image
2K6 credits / image$0.030 / image
4K10 credits / image$0.050 / image

Best Use Cases

  • Commercial posters and infographicsUse multilingual text rendering and mixed text and image layouts to quickly create advertising posters, promotional brochures, and chart illustrations with clearly arranged slogans.

  • Ultrawide wallpapers and panoramic concept scenesGenerate expansive landscapes, science fiction cities, and film storyboard concepts in wide formats such as 21:9 and 16:9 without tiling seams.

  • E-commerce product promotions and concept visualsExplore product presentation concepts in batches with different styles, material textures, and lighting moods, substantially shortening visual planning time.

  • Visual asset production for multiple social channelsUse low costs and fast responses to produce consistently styled feed visuals in suitable sizes for multiple social media accounts.

Pro Tips

  • Put display text in double quotation marks: Wrap English or multilingual phrases that must appear accurately in straight double quotes, such as "SPRING FESTIVAL 2026", to improve letter edge sharpness and spelling accuracy.
  • Specify spatial depth and viewpoint: Include camera angles such as a 45-degree downward view or a close-up, along with lighting descriptions such as cool, diffuse morning light or a sunset backlit silhouette, to make the most of the model's spatial capabilities.
  • Use wide formats to avoid repeating textures: Choose 21:9 or 16:9 directly when you need a wide composition. The new architecture fixes tiling artifacts, removing the need for manual cropping and stitching.
  • Match resolution to the task: Use the 4-credit 1K mode for initial concepts and batch exploration. Once you choose a direction, generate the final image in 2K or 4K to balance efficiency, detail, and cost.

Notes

  • prompt is required and supports up to 20,000 characters for detailed descriptions of the scene and any text requirements.
  • The Text to Image endpoint generates images from text alone. Use Nano Banana 2.1 Edit to modify existing assets or combine visual references.
  • Resolution determines the final charge. Each request produces one image, costing 4 / 6 / 10 credits for 1K / 2K / 4K respectively. Confirm the required output settings before submitting.
  • Tasks run asynchronously. Submission returns a task_id; once processing finishes, the status response contains the finished image URL with an embedded SynthID digital watermark.

Nano Banana 2.1 Text to Image API frequently asked questions

What is the Nano Banana 2.1 Text to Image API?

Nano Banana 2.1 Text to Image is Google's efficient model for generating images from text. It produces native 1K, 2K, and 4K images from text prompts, with strong multilingual text layout rendering and panoramic output free of tiling artifacts. Built on Gemini 3.6 Flash's native multimodal reasoning architecture with a built-in thinking mechanism, it follows complex perspective and lighting descriptions while delivering commercial image quality at an accessible cost and high speed. You can call it programmatically through the API or try it in the playground above.

Can Nano Banana 2.1 Text to Image generate text within images?

Yes. Its enhanced text layout rendering engine accurately generates short multilingual text for poster headlines, packaging, signs, and chart labels. Enclosing specific words or phrases in straight double quotation marks within the prompt can further improve spelling accuracy and edge clarity.

Which output resolutions and aspect ratios does Nano Banana 2.1 Text to Image support?

The model natively supports 1K, 2K, and 4K resolutions and 10 aspect ratios: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, and 21:9. Choose a standard square, portrait poster, or cinematic ultrawide format to match your publishing channel.

How does Nano Banana 2.1 Text to Image address tiling artifacts in panoramic images?

Earlier models could repeat image content or produce tile-like defects in wide formats. Nano Banana 2.1 reoptimizes the attention mechanism for extreme aspect ratios, maintaining overall coherence and natural continuity in 2K and 4K panoramic and wide-format generation.

How does Nano Banana 2.1 compare with Nano Banana 2 and Nano Banana Pro?

Nano Banana 2.1 is a major upgrade to Nano Banana 2, retaining Flash generation speed while substantially improving image quality, text generation, and composition, with an approximately 50% reduction in cost per image. Compared with the top-tier Nano Banana Pro, version 2.1 is more cost-effective for everyday, frequent generation at scale.

How much does Nano Banana 2.1 Text to Image cost?

Pricing depends on the resolution value in the request. Each 1K image costs 4 credits ($0.020), 2K costs 6 credits ($0.030), and 4K costs 10 credits ($0.050). Each credit is worth $0.005.

How should I write prompts for the best Nano Banana 2.1 Text to Image results?

Structure your prompt around specific subject appearance, material properties, focal length, camera angle, and environmental light direction. If the image includes text, mark it precisely with double quotation marks. Avoid vague, abstract superlatives.