Nano Banana 2.1: Features, Benchmarks, Pricing, and Availability
Introduction
Google's Nano Banana 2.1 is an updated image generation and editing model designed to improve image quality, prompt accuracy, text rendering, and subject consistency.
Compared with Nano Banana 2, the newer version focuses on making image editing more reliable and improving infographic creation. It also introduces lower image-generation prices at 1K, 2K, and 4K resolutions, although input and reasoning costs are higher.
What Is Nano Banana 2.1?
Nano Banana 2.1 is Google's latest version of its Nano Banana image model. Its API model ID is gemini-nano-banana-2.1.
Nano Banana 2 is also known as Gemini 3.1 Flash Image, while Nano Banana Pro is a separate model in the same family.
Google has deprecated Nano Banana 2, but it remains available without an announced shutdown date. Businesses using the older version should evaluate compatibility before migrating existing workflows.
What Are the New Features of Nano Banana 2.1?
Nano Banana 2.1 focuses on improving realism, following detailed prompts, and generating images at different resolutions, including 1K, 2K, and 4K. These improvements can be useful for marketing visuals, product imagery, social media graphics, and other creative projects. However, small text can still appear blurry, and precise left-right placement remains a potential limitation. |
The model is designed to make targeted edits while preserving other elements of an image. For example, users can change an object's color, adjust a background, or modify a scene while keeping the main subjects recognizable. This is particularly useful for workflows that require several revisions of the same image. |
Nano Banana 2.1 aims to produce clearer text and more structured infographic layouts. These capabilities can help users create educational diagrams, presentation graphics, process illustrations, and marketing materials. Nevertheless, generated text and complex layouts should be reviewed before publication. |
4. Features Carried Over From Nano Banana 2Nano Banana 2.1 retains several capabilities from the previous version:
The newer version also addresses tiling issues in wide images and removes the 512-pixel output option. |
Nano Banana 2.1 Benchmarks
Google reports improvements in overall image preference, infographic design, and general editing compared with Nano Banana 2.
Benchmark | Nano Banana 2.1 | Nano Banana 2 |
Overall preference (Elo) | 1050 ± 14 | 990 ± 7 |
Infographic design (Elo) | 1048 ± 17 | 961 ± 12 |
Infographic factuality (AutoRater) | 0.521 | 0.179 |
General editing (Elo) | 1026 ± 12 | 938 ± 11 |
The Elo scores are based on comparative human preferences, while infographic factuality uses a separate automated evaluation metric. These measurements should not be compared directly with one another.
The results suggest that Nano Banana 2.1 has improved in infographic creation and image editing. However, benchmark results do not guarantee identical performance across every prompt or creative task.
Nano Banana 2.1 Pricing
Google uses token-based pricing for its Gemini API rather than charging one flat fee for every generated image.
According to the supplied pricing figures, Nano Banana 2.1 reduces image-output costs while increasing input and reasoning rates.
API Pricing Comparison
Pricing category | Nano Banana 2.1 | Nano Banana 2 |
Input, per 1M tokens | $1.50 | $0.50 |
Text and reasoning output, per 1M tokens | $7.50 | $3.00 |
1K image | $0.0336 | $0.067 |
2K image | $0.0504 | $0.101 |
4K image | $0.113 | $0.151 |
At 1K and 2K resolutions, image generation costs approximately half as much as Nano Banana 2. At 4K, the reduction is around 25%.
However, input costs are three times higher, and text and reasoning output costs are 2.5 times higher. As a result, workflows involving long prompts or multiple reference images may see smaller overall savings.
Google's listed API pricing does not include a free API tier for either model.
Nano Banana 2.1 Availability
Nano Banana 2.1 is available through several Google platforms, although subscription access and API billing operate separately.
Gemini app: Access depends on the supported Google AI subscription plan and account availability.
Google Flow: Nano Banana 2.1 is included among the available image-generation models.
Google AI Studio: Developers can experiment with the model and access it through supported configurations.
Gemini API: Developers can integrate image generation and editing into applications using the model ID gemini-nano-banana-2.1.
Availability and usage limits may vary by account, subscription, region, and platform. API usage is billed separately from consumer subscriptions.
When Should You Use Nano Banana 2.1?
Nano Banana 2.1 is worth considering for:
Marketing and social media: Creating promotional images and branded visual content.
Image editing: Making targeted changes while preserving the main subjects.
Infographics: Producing diagrams, educational visuals, and process illustrations.
Reference-based generation: Using multiple reference images to guide the output.
High-resolution creative work: Generating images at 2K and 4K resolutions.
Nano Banana 2 may remain relevant for existing workflows that depend on 512-pixel output or benefit from its lower input and reasoning costs. Since Google has deprecated it, teams should consider testing and planning a migration to the newer version.
Conclusion
Nano Banana 2.1 focuses on improving image generation, editing, text rendering, and infographic creation. Google's reported benchmarks show improvements over Nano Banana 2, while lower image-output prices make the updated model attractive for certain creative workflows.
However, higher input and reasoning costs, along with limitations in small-text rendering and precise object placement, should be considered before adopting it at scale.
For businesses and creators, the best approach is to evaluate image quality, editing requirements, and total API costs against their specific needs.
Comments