Google has officially unveiled its anticipated AI image editor, known as Nano Banana, confirming weeks of speculation among technology enthusiasts. The company announced that this uniquely named tool is now available as Gemini 2.5 Flash Image, integrated directly into the Gemini app for users globally. The excitement surrounding Nano Banana began when peculiar yet consistently edited AI photos started appearing online, generating curiosity about the underlying technology. Last evening, Google CEO Sundar Pichai addressed the speculation with a playful post on X (formerly Twitter). His message, featuring three banana emojis, was accompanied by a collection of images showing his dog, Jeffree, reimagined in various roles such as a surfer, cowboy, superhero, and chef.
The edits, shared in celebration of International Dog Day, showcased the model’s capability to preserve Jeffree’s unique appearance across different contexts. Pichai wrote: “Our image editing model is now rolling out in @Geminiapp – and yes, it’s (banana emojis). Top of @lmarena’s image edit leaderboard, it’s especially good at maintaining likeness across different contexts. Check out a few of my dog Jeffree in honour of International Dog Day – though don’t let these fool you, he definitely prefers the couch:).” The Gemini 2.5 Flash Image model enhances Google’s prior ventures into generative visuals. A common criticism of AI image generation has been its struggle to retain likeness—whether for pets or people.
Nano Banana aims to tackle this issue by centering edits around the original subject’s distinctive features. The tool is designed for ease of use: users upload a photo, input a prompt, and the AI applies edits that honor the subject’s characteristics. In addition to single-step edits, Nano Banana introduces multi-turn editing, allowing users to refine or expand upon images incrementally. Another feature is design mixing, which permits textures and patterns from one image to transfer to another. For example, butterfly wings could inspire a dress print, or rain boots could feature the texture of flower petals. The model also facilitates photo blending, enabling multiple images to merge into a single composition seamlessly.
Edited images can even be reintroduced into Gemini to create animated video outputs, broadening creative avenues. The rollout of Nano Banana is comprehensive, as Google confirmed it is accessible worldwide through the Gemini app for both free and paid users. To uphold transparency, every AI-generated image will include Google’s SynthID digital watermark, both visible and invisible, clearly differentiating them from authentic photographs. For businesses and developers, the model is also available via the Gemini API, Google AI Studio, and Vertex AI, with pricing set at $30 per million output tokens, translating to approximately $0.039 per image generated.
With its playful branding and advanced features, Nano Banana embodies Google’s ongoing effort to make AI both accessible and creative, while addressing a long-standing challenge in image generation: ensuring that edits maintain a personal and authentic touch.