Google launches Nano Banana 2 that combines professional image reasoning with “ultra-fast” speed

Google has launched its new image generation model with artificial intelligence (AI) Nano Banana 2 (Gemini 3.1 Flash Image), which offers reasoning capabilities and high precision to generate professional images, but with “ultra-fast” speed.

After the launch of Nano Banana Pro in November of last year, which allows generating images with maximum factual precision, the company has presented its new model, in this case, focused on offering faster results, but maintaining precise following of the instructions.

This is Nano Banana 2, that is, Gemini 3.1 Flash Image, its latest generation model, which has been designed to dramatically reduce the gap between speed and visual fidelitywith results that improve the visual fidelitywith more vibrant lighting, more varied textures and sharper details, as well as quality realistic up to 4K.

This was announced by the technology giant in a statement on its blog, where it assured that Nano Banana 2 carries “the Gemini Flash high-speed intelligence to visual generation“, allowing you to carry out photo edits or generate images quickly.

Specifically, Google has explained that, to generate its images, the new model extracts information from the database. Gemini real world knowledge and is based on real-time information from the Search in the weballowing you to represent topics more accurately.

Likewise, it also offers high capacities when it comes to render images and translate text accurately, so that it allows you to generate precise and legible text for marketing mockups or greeting cards, for example. Even locating and translating text within a photograph.

Continuing with the capabilities of Nano Banana 2, Google has claimed that it has improved subject consistency, as the model can maintain the likeness of up to five characters and the fidelity of up to 14 objects in a single workflow. This allows different narratives to be created without altering the appearance of different results.

Likewise, Google ensures a precise following of instructions, That is, the model strictly adheres to the users’ requests, capturing the nuances of the idea so that the generated image is as faithful as possible to what was requested.

Furthermore, Nano Banana 2 can also generate images with various aspect ratios and resolutions, based on users’ needs, from 512 px to 4K.

It must be taken into account that, hand in hand with this model, Google has stressed that they continue to promote their SynthID watermarking technology, along with C2PA content credentials, in images with AI. In this way, they inform users not only if AI has been used, but also how it has been done. In this regard, he announced that C2PA verification will soon be incorporated into the Gemini application.

With all this, Google has indicated that Nano Banana 2 is now available on Google products, including the Gemini app, where it will replace Nano Banana Pro on Quick, Thinking, and Pro models (except for Google AI Pro and Ultra subscribers, who will continue to have access for specialized tasks).

The new model will also be available in search, through AI Mode and Lens, in the Google application and in mobile and desktop browsers. Likewise, it can be found in AI Studio + API, in Google Cloud, Flow and Google Ads.