Background

Google’s Nano Banana quickly became a fan favorite after its debut last fall, impressing users with photorealistic capabilities that outpaced competitors such as Midjourney and OpenAI. The tool’s ability to edit existing photos as well as generate new images set it apart, and a Pro version later expanded on those strengths with improved text generation.

New Features in Nano Banana 2

The second‑generation Nano Banana builds on the original’s speed while inheriting the detailed accuracy of the Pro model. Powered by the world knowledge embedded in Gemini 3, the new base model is designed to stay closer to reality and reduce hallucinations. Google says the update delivers “vibrant lighting, richer textures and sharper details,” and it can now generate images up to 4K resolution in a variety of aspect ratios.

Text rendering within images—such as for greeting cards—has been refined for greater clarity, and character consistency receives a boost, addressing creator feedback that highlighted Nano Banana’s strength in that area. Users will find Nano Banana 2 automatically replacing the original model across Gemini, with the option to select the Pro alternative if desired.

Addressing AI‑Generated Content Concerns

Since Nano Banana’s launch, the proliferation of AI‑generated media has sparked debate about its impact on online ecosystems. Low‑quality AI content, sometimes called “slop,” now floods social platforms, and a CNET survey found that while the majority of users believe they see AI‑generated posts, only 44% feel confident they can identify them.

To help combat this uncertainty, Google is attaching invisible watermarks—referred to as content credentials—to images created with Nano Banana 2. When an image is uploaded to Gemini, the system can determine whether it originated from Google’s AI, offering a method for verification, though it works only for images generated by Google’s tools.

Used: News Factory APP - news discovery and automation - ChatGPT for Business