Google launches Nano Banana 2 Lite with 4-second AI image generation capabilities

Google has introduced Nano Banana 2 Lite, the latest and fastest model in its Nano Banana image generation family, with an image generation time of four seconds. The company is positioning it as a fast image-generation model for developers handling high volumes of images.

Alongside the new image model, Google has expanded access to Gemini Omni Flash, its multimodal model for video generation and conversational editing. The company said the two models are designed to work together, enabling developers to build end-to-end AI workflows that move from image generation to short video creation.

Must read: Google caps Meta's Gemini AI usage over compute limits

Google expands image and video AI tools for developers

Nano Banana 2 Lite, listed as gemini-3.1-flash-lite-image, is aimed at high-speed creative workflows where latency and operating costs matter more than advanced reasoning. Google said it is suited for rapid prototyping, visual brainstorming, large-batch image generation and other high-volume use cases. The company has also recommended it as the replacement for the original Nano Banana model, gemini-2.5-flash-image.

Google has priced Nano Banana 2 Lite at $0.034 per 1K-resolution image, making it the lowest-cost option in the Nano Banana family for developers managing large-scale workloads or tighter budgets. Despite its focus on speed, Google said the model delivers reliable prompt adherence, consistent characters across generated images and readable text within images. 

Must read: Google predicts Venezuela earthquake, but there’s a catch; Here's the science behind Alert

The company said Nano Banana 2 Lite is built for ultra-fast, high-volume generation, while Nano Banana 2, or Gemini 3.1 Flash Image, remains its all-round model. At the premium end, Nano Banana Pro, or Gemini 3 Pro Image, is meant for workloads requiring greater reasoning, precision and creative control.

Google said the model is available through Google AI Studio, the Gemini API and the Gemini Enterprise Agent Platform. It is also being rolled out across consumer services, including AI Mode in Search, the Gemini app, NotebookLM, Google Photos, Google Flow, Stitch and Google Ads.

Gemini Omni Flash, first unveiled at Google I/O, is now available to developers in public preview through Google AI Studio and the Gemini API. Google said the model supports high-quality video generation and conversational editing using text, images and video as inputs. It is designed to let users refine videos through natural language prompts while maintaining consistency across multiple edits, and it supports multimodal referencing to guide results more accurately. The model is priced at $0.10 per second of generated video, the same as Veo 3.1 Fast.

For now, Omni Flash generates videos of up to 10 seconds, with support for longer videos planned in a future update. Audio reference uploads, scene extension and reliable processing of short video references are still under development. Google said developers can use Nano Banana 2 Lite to generate images first and then pass those images to Gemini Omni Flash to turn them into short videos.