The headline changes are image quality, prompt accuracy, text rendering, and subject consistency. Nano Banana 2 already did all four, so I wanted to see whether the update was visible in normal use. To that end, I’m giving both models the same four prompts for generation and editing, and I’ll render a verdict.<\/p>\n
TL;DR<\/h2>\n\n- Nano Banana 2 still works but is now deprecated with no announced shutdown date.<\/li>\n
- Google reports higher scores for 2.1 in overall preference, infographics, and editing, while small text and left and right placement remain weak spots.<\/li>\n
- Image output costs roughly half as much at 1K and 2K and about 25% less at 4K, but input and reasoning rates are higher.<\/li>\n
- The difference matters most for repeated edits and layouts with text. A simple first image may look similar on either model.<\/li>\n<\/ul>\n
What Is Nano Banana 2.1?<\/h2>\n
Nano Banana 2.1 is the current version of Google’s Nano Banana image model. Its application programming interface (API)<\/a> ID is gemini-nano-banana-2.1<\/code><\/a>.<\/p>\nThe naming is the confusing part. Nano Banana 2 is also called Gemini 3.1 Flash Image, while Nano Banana Pro<\/a> is a separate model in the same family. Nano Banana 2 still works, but Google has deprecated it without announcing a shutdown date<\/a>.<\/p>\nWhat Are the New Features of Nano Banana 2.1?<\/h2>\n
Version 2.1 focuses on reliability rather than new tools.<\/p>\n
Image quality, prompt accuracy, and text rendering<\/h3>\n
Google reports changes to realism at 1K, 2K, and 4K, prompt accuracy, text, and infographic layouts. Its model card<\/a> still lists blurry small text and errors with left and right.<\/p>\nImage editing and subject consistency<\/h3>\n
Google lists subject consistency as another change in 2.1. Here, that means changing one element without altering the character, objects, or scene around it.<\/p>\n
Features carried over from Nano Banana 2<\/h3>\n
Nano Banana 2 already supported:<\/p>\n
\n- Up to 14 reference images<\/li>\n
- Google Search grounding<\/li>\n
- 4K output<\/li>\n
- Very wide aspect ratios such as 8:1<\/li>\n<\/ul>\n
Version 2.1 fixes tiling in wide images and drops the 512px option.<\/p>\n
Nano Banana 2.1 Benchmarks<\/h2>\n
Google reports two types of scores. Overall preference, infographic design, and general editing come from human side-by-side comparisons converted into Elo; higher means raters preferred the model more often. Infographic factuality uses a separate AutoRater metric, so its decimal score is not Elo.<\/p>\n
The card labels both models as “Thinking” without naming minimal<\/code>, medium<\/code>, or high<\/code>. That does not establish that Google used the same high<\/code> setting as my hands-on test.<\/p>\nIf you only care about the images, skip this table and go to the hands-on test.<\/p>\n
\n\n
What Is Nano Banana 2.1?<\/h2>\n
Nano Banana 2.1 is the current version of Google’s Nano Banana image model. Its application programming interface (API)<\/a> ID is The naming is the confusing part. Nano Banana 2 is also called Gemini 3.1 Flash Image, while Nano Banana Pro<\/a> is a separate model in the same family. Nano Banana 2 still works, but Google has deprecated it without announcing a shutdown date<\/a>.<\/p>\n Version 2.1 focuses on reliability rather than new tools.<\/p>\n Google reports changes to realism at 1K, 2K, and 4K, prompt accuracy, text, and infographic layouts. Its model card<\/a> still lists blurry small text and errors with left and right.<\/p>\n Google lists subject consistency as another change in 2.1. Here, that means changing one element without altering the character, objects, or scene around it.<\/p>\n Nano Banana 2 already supported:<\/p>\n Version 2.1 fixes tiling in wide images and drops the 512px option.<\/p>\n Google reports two types of scores. Overall preference, infographic design, and general editing come from human side-by-side comparisons converted into Elo; higher means raters preferred the model more often. Infographic factuality uses a separate AutoRater metric, so its decimal score is not Elo.<\/p>\n The card labels both models as “Thinking” without naming If you only care about the images, skip this table and go to the hands-on test.<\/p>\ngemini-nano-banana-2.1<\/code><\/a>.<\/p>\nWhat Are the New Features of Nano Banana 2.1?<\/h2>\n
Image quality, prompt accuracy, and text rendering<\/h3>\n
Image editing and subject consistency<\/h3>\n
Features carried over from Nano Banana 2<\/h3>\n
\n
Nano Banana 2.1 Benchmarks<\/h2>\n
minimal<\/code>, medium<\/code>, or high<\/code>. That does not establish that Google used the same high<\/code> setting as my hands-on test.<\/p>\n
![](\"https:\/\/media.datacamp.com\/cms\/2fb06ca37f00fdada8243aa743976490.png\")
<\/p>\n![](\"https:\/\/media.datacamp.com\/cms\/a87d85dda11c9b3dc48dced7f91dfaa6.png\")
<\/p>\n![](\"https:\/\/media.datacamp.com\/cms\/ab3c8024a721941797edbc6280d79bfe.png\")
<\/p>\n![](\"https:\/\/media.datacamp.com\/cms\/ecf5a65ca592a64bf5391739d6c789df.png\")
<\/p>\n