What Is Nano Banana 2 Lite? The 4-Second, $0.034 Image AI

What is Nano Banana 2 Lite

Nano Banana 2 Lite is an image generation and editing model released by Google on June 30, 2026. As the name suggests, it is a lightweight take on the Nano Banana line, prioritizing speed and price. For the bigger picture on Gemini and its higher-end models, see our guide to Gemini's latest models (Gemini 3.5).

A low-cost design: about 4 seconds, $0.034 per image

The defining traits of Nano Banana 2 Lite are speed and price. It generates an image from text in about 4 seconds, at $0.034 per 1K-resolution image. Google calls it the fastest and most cost-efficient model in the Nano Banana family.

Note that $0.034 is per single 1K-resolution image (around 1024 pixels), not per 1,000 images. For workflows that churn through drafts and ideas, that low per-image cost adds up in your favor.

Prompt fidelity, character consistency, legible text

Lightweight models often lose quality, but Nano Banana 2 Lite emphasizes practical polish. It keeps prompt fidelity, character consistency across repeated renders, and legibility of text placed inside images.

That makes it easy to use where you need volume — banner images, social drafts, and character-based materials.

The video AI: Gemini Omni Flash

Alongside Nano Banana 2 Lite, Google announced a model on the video side. Gemini Omni Flash stands out for producing video cheaply and refining it through natural-language conversation.

10-second video at $0.10 per second

Gemini Omni Flash is a video generation model rolling out to developers. It currently produces 10-second videos at $0.10 per second of output. Google says this rate matches the existing Veo 3.1 Fast.

That works out to roughly $1 for a 10-second clip. It suits short ad drafts and social videos you want to try in volume while watching costs.

Accepts text, image, and video inputs

Omni Flash takes not just text but also images and video as inputs. Combined with conversational editing, you can refine results with natural instructions like "animate this image" or "make it a bit brighter." Since it is still in preview, features and limits may change.

Availability and how to use the two models

Finally, here is where you can use the two models and how to pick between them.

Availability (AI Studio, Gemini API, Enterprise Agent Platform)

Nano Banana 2 Lite is available in Google AI Studio, the Gemini API, and the Gemini Enterprise Agent Platform, and is rolling out to consumer surfaces such as AI Mode in Search and the Gemini app. Gemini Omni Flash is available in public preview in Google AI Studio and the Gemini API.

Chaining image to video

The models work on their own, but they shine when connected. You can pass an image made with Nano Banana 2 Lite straight to Gemini Omni Flash as a reference, producing a single flow from still image to video.

In short: use Nano Banana 2 Lite to make stills in volume at low cost, then Gemini Omni Flash to turn them into short videos. Google's same-day announcements also included the desktop-automation agent in our guide to Gemini Spark on macOS; read together, they show Google pushing AI on both the app and base-model fronts.

When gathering prompt sources or reference material from the web, converting pages to Markdown first keeps their structure intact and makes them easier for an AI to handle.

Free ToolURL to Markdown ConverterConvert any public web page URL to Markdown. Preserves headings, tables, lists, and links — perfect for LLM and RAG preprocessing, research notes, and archiving web articles.Try it now →