Google: Nano Banana 2 (Gemini 3.1 Flash Image)

Google Nano Banana 2 is a multimodal model from Google that accepts both text and image inputs, with a context window of 131,072 tokens and a maximum output of 32,768 tokens. It supports reasoning, but does not support tool use, and structured output support is unconfirmed. That combination makes it suited to tasks where visual understanding and extended context matter, but not to agentic or tool-calling workflows. At $0.50 per million input tokens and $3.00 per million output tokens, the pricing sits on the accessible end for input while output costs are moderate. The more significant limitation for buyers is that there is no independent benchmark coverage available, so performance relative to competing models is unproven. Teams willing to run their own evaluations may find it worth testing for image-plus-text tasks, but those who need validated performance data before committing should wait for third-party results.