Google: Nano Banana (Gemini 2.5 Flash Image π) API
The dedicated image model of the Gemini 2.5 family, also known as Nano Banana. It is the first commercial model to combine multi-turn conversational editing with robust character consistency. It functions as an "AI Visual Designer" capable of complex reasoning and seamless multi-image fusion.
- Context window: 32,768 tokens
- Max output: 32,768 tokens
- Input: text, image
- Output: image
- Reasoning: Supported
- File input: Supported
- Released: 2025-08-26
- Knowledge cutoff: 2025-06-01
Frequently Asked Questions
What is the context window of Nano Banana (Gemini 2.5 Flash Image π)?
Nano Banana (Gemini 2.5 Flash Image π) supports a context window of up to 32,768 tokens.
Does Nano Banana (Gemini 2.5 Flash Image π) support reasoning?
Yes. Nano Banana (Gemini 2.5 Flash Image π) is a reasoning-capable model.
What is the knowledge cutoff of Nano Banana (Gemini 2.5 Flash Image π)?
The knowledge cutoff of Nano Banana (Gemini 2.5 Flash Image π) is 2025-06-01.