On Tuesday local time, tech giant Google officially released its new self-developed AI image and video generation model Nano Banana 2 Lite. The official stated that compared to the previous standard version, the new version has achieved significant improvements in operating speed while greatly reducing user costs.

The most core advantage of this model is its extremely low latency, which can generate a high-quality image in just 4 seconds. This feature makes it particularly suitable for professional workflows that require repeated rapid modifications or face large-scale image production demands within a short period of time.

image.png

Available across all channels and replacing the old version

In terms of pricing, generating an output image of 1K resolution requires 1120 tokens, with a billing cost of 0.034 dollars (approximately 0.23 Chinese yuan). Currently, the model has been officially launched on Google AI Studio, Gemini API, and the Gemini enterprise intelligent platform, and will completely replace the first-generation model classified as "old version".

In addition to the image generation model, Google has also expanded the open scope of Gemini Omni Flash, with a video generation billing rate of 0.10 dollars per second. At the same time, the new demo application Omni Product Studio can convert static images into "movie-level" e-commerce videos with one click, helping developers build an end-to-end multimedia creation experience.