Google might be ready to serve another round of its “Nano Banana.” New hints across Gemini’s interface suggest a sequel to the viral AI image tool, rumored to be called Nano Banana 2 (or GEMPIX 2), could be about to launch.
Developers and AI watchers first noticed references to Nano Banana 2 in Gemini’s web interface, specifically through new “announcement cards” signaling an upcoming release. Such pre-release signals have historically appeared shortly before Google’s public rollouts.
If Google follows its usual pattern, that could mean Nano Banana 2 arrives in mid-November, alongside the rumored Gemini 3 Pro preview.
What could be coming
The original Nano Banana, built on Google’s Gemini 2.5 Flash Image, became a social media phenomenon earlier this year. Users turned photos into cinematic, 3D-like portraits that flooded Instagram and X, creating a viral wave that pulled millions of new users into the Gemini app within just a few weeks.
Now, its successor, Nano Banana 2, appears to be a more advanced, professional-grade model. Several reports have started painting a picture of what the next Nano Banana might deliver. The upgraded model is said to focus heavily on quality, speed, and precision.
Early leaks suggest it could produce 4K-ready visuals with much sharper fidelity and give users more control over how images are generated and edited. Performance is also a major focus. The model reportedly handles complex prompts in under 10 seconds, making image creation almost instant.
Additionally, sources report improvements in character consistency and text accuracy, addressing one of the most common frustrations from the initial version.
There are also hints of a new “Edit with Gemini” tool, allowing users to fine-tune specific parts of an image without having to start from scratch. To round it off, Google is reportedly enhancing professional safeguards, such as built-in metadata and watermarking, to ensure the responsible use of AI-generated content.
Built on Gemini 3 pro image
According to Testing Catalog, code for a “Gemini 3 Pro preview model” labeled “gemini-3-pro-preview-11-2025” has appeared in the VertexAI developer platform, signaling a launch window that aligns with earlier rumors of a November reveal. This new flagship model is speculated to feature an impressive 1 million token context window, making it ideal for processing vast amounts of data.
There is “ongoing speculation about whether Nano Banana 2 will leverage Gemini 3 Pro under the hood, or if the two will be rolled out together as part of a larger multimodal update across Google’s AI stack,” Testing Catalog noted.
So far, Google hasn’t made any official statement about Nano Banana 2 or its timing. The company typically keeps AI releases under wraps until the last minute, often seeding hints in its code or UI just before launch. That means while excitement is mounting, everything remains speculative until Google confirms it.
And don’t miss our latest scoop: Google just pulled its open-source model Gemma after a defamation claim from Marsha Blackburn hit the headlines — significant tremors in the AI ethics landscape.