Nano Banana 2 Lite
Neural Network
Nano Banana 2 Lite is a Google text-to-image model, available via API in BotHub.
Specifications
- Input
- text, images
- Output
- images
- Aspect ratio
- 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 16:9, 9:16, 21:9
- Capabilities
- Batch API
Image prompt
(per 1K tokens)
How it works Nano Banana 2 Lite?
Nano Banana 2 Lite is a compact model from Google's Gemini family, focused on speed and cost-efficiency: it creates images from text descriptions and is designed for high-frequency requests when you need to iterate through dozens of visual options in a single session. This format is suitable where the pace and cost of experimentation are more important than long work on a single frame. In BotHub, the model works without a VPN or foreign card: you pay with a Russian card as you go, for consumed tokens, and the unused balance does not expire. Nearby in the same interface are over 250 neural networks, so you can assemble a draft here and send the final frame to a heavier model without changing windows. For product tasks, there is a unified OpenAI-compatible API, AES-GCM encryption, and corporate access with a contract, invoice, and admin panel. The model helps designers quickly sketch mood boards and concepts, marketers prepare creative options for tests, developers integrate image generation into pipelines, and authors of articles and presentations get illustrations for specific material.Frequently asked questions about Nano Banana 2 Lite
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
nano-banana-2-lite is a Google image generation model: it creates an image from a text description and modifies an existing one based on a reference. It is suitable for illustrations, concepts, post visuals, and product cards, and the batch mode helps complete series of similar images in one pass.
Specific resolution and file formats are not specified in our data, so focus on the result: the model accepts a description or source image and returns a finished picture. Context is 65,536 tokens, output limit is 32,000 tokens, which is enough for a detailed prompt and a series of frames.
A separate reasoning mode before answering is not noted in our data. However, the model belongs to the text-to-image LLM class and works with detailed descriptions, so the large context allows you to set the scene, style, and details with one detailed prompt.
Function calling and structured JSON output are not noted for this model in our data — it is an image generator, not a data retrieval tool. If you need strict JSON, connect a text model in the same BotHub OpenAI-compatible API without rewriting the integration.
Images — yes: the image-to-image mode allows you to upload a reference and refine it, change the style, background, or individual elements. Audio and video input is not noted in our data, so for such files, choose a model with the required modality in the BotHub catalog.
We have no data on prompt languages, so it is safer to test with several phrasings and compare the result. The service itself is fully Russian-speaking: the website and Telegram bot, payment with Russian cards in rubles, access without VPN and foreign cards, pay-as-you-go for tokens.