Nano Banana 2.1
GeminiSpecialized
ThinkingTool UseVisionStructured Output
About this model
Nano Banana 2.1 (Gemini Nano Banana 2.1) is Google's image generation and editing model on the Flash tier, succeeding Nano Banana 2 and Nano Banana Pro. It improves product recontextualization,...
Performance Tier
Specialized
Nano Banana 2.1 is a specialized model from Gemini : built for a specific domain.
Domain-specific model. Optimized for a particular task such as code generation, image creation, or web search.
Pricing
This model is included in Elosia plansModerate
Moderate cost. A balanced choice for regular use.
| Type | per 1M tokens |
|---|---|
| Input (prompt) | $1.50 |
| Output (completion) | $7.50 |
Capabilities
Context Length66K
Max Output Tokens59K
TokenizerGemini
Inputimage, text
Outputimage, text
Release DateOctober 6, 2026
Recommended Use Cases
Creative Writing
Strengths
- Image generation and conversational editing at Flash-level speed, with output at 1K, 2K and 4K resolution
- Accurate text rendering and infographic layout, including localized text in several languages
- Multi-image fusion from up to 14 reference images, keeping up to 4 characters and 10 objects consistent
- Mask-based editing and multi-turn editing for precise, iterative changes
- Wide and panoramic aspect ratios from 1:8 to 8:1
Limitations
- Small text, long paragraphs and full-page text often render poorly, especially at 1K resolution
- Character consistency between input and output images is not always perfect, and masked or doodle-based edits can follow instructions only partially
- Occasional confusion with spatial positions such as left and right, and limited 3D reasoning and factual accuracy
- Specialized for image generation and editing, not general reasoning
Frequently asked questions
Resources
This model may use your data for training
