Do you want smarter insights in your inbox? Sign up for our weekly newsletters to get what is vital for AI leaders, data and security. Subscribe now
Google The Flash Gemini 2.5 image published, a modern model that many beta users knew as Nanobanana, which gives enterprises a greater selection of inventive projects. This allows them to quickly change the images they need, with more control than what previous models were offered.
The model will be integrated with the Gemini application.
The model, built on Gemini 2.5 Flash, adds more possibilities to the native editing of the image in the Gemini application. The image of Flash Gemini 2.5 maintains the similarity of the character between different images and has more consistency when editing photos. If the user sends a photo of his pet and then asks the model to change the background or add a hat to the dog, the image of Flash Gemini 2.5 will do it without changing the subject of the image.
“We know that when editing photos of yourself or people you know well, subtle defects of matter, the performance” close, but not quite the same “does not seem right,” said Google on a post on a blog written by Multimodal Generation David Sharon and Google Deepmind Gemini Product Product Product Product Nicole. Homemade look consistently like them. ”
AI scaling hits its limits
Power capitals, the growing costs of the token and inference delay are transforming AI Enterprise. Join our exclusive salon to discover how the best teams are:
- Changing energy into a strategic advantage
- Architect of effective inference regarding real capacity profits
- Unlocking competitive roi using balanced AI systems
Secure your place to remain ahead: https://bit.ly/4mwgni
One of the complaint companies and some individual users is that when installing the editions in the paintings generated by AI, compact amendments change the photo too much. For example, someone can instruct the model to move the position of the person in the photo, and although the model does what is said, the face of the person is slightly changed.
All paintings generated on the twins will contain the Google synthesis. The model is available to all paid and free Gemini application users.
Speculation that Google plans to release a modern picture model raging on social platforms. Users at LM Arena saw how the mysterious modern model called Nanobanana, which took place after “complex, multi -stage instructions with impressive accuracy”, such as Andressen Horowitz’s partner Justine Moore put it in the post.
People soon noticed that the Nanobanana model seemed to come from Google before confirming several early testers. Although at that time Google did not confirm what he planned to do with the LM Arena model.
Until this week, speculation about when a model will appear, which is in a sense prophetic.
Many emotions come as a fight between models suppliers to offer more talented and realistic images and editions, showing how powerful multimodal models have become.
However, Google still has to fight rivals like QWEN and his recently released edition Qwen-Image i Openaiwhich added the native edition of AI images to ChatgPT, and also made available the model as API.
Of course, AdobeHe has long been considered one of the leaders in the space of image editing, he added his flagship Firefly model to Photoshop and other photo editing platforms.
Edition of the native image
Gemini added the native edition of AI paintings in Gemini in March, which he offered free users of the chat platform.
The introduction of the image editing function directly to the chat platform would allow enterprises to repair images or charts without carrying Windows.
Users can send a photo to Gemini, and then inform the model what changes want. After their satisfaction, modern photos can be again sent to Gemini and transformed into video.
In addition to adding a costume or changing the location, the Flash Gemini 2.5 image can combine different photos, offers the editing of many inverting and mixing styles of one photo.
