{"id":177327,"date":"2025-08-26T15:19:16","date_gmt":"2025-08-26T15:19:16","guid":{"rendered":"https:\/\/www.europesays.com\/us\/177327\/"},"modified":"2025-08-26T15:19:16","modified_gmt":"2025-08-26T15:19:16","slug":"google-geminis-ai-image-model-gets-a-bananas-upgrade","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/us\/177327\/","title":{"rendered":"Google Gemini’s AI image model gets a ‘bananas’ upgrade"},"content":{"rendered":"

Google is upgrading its Gemini chatbot with a new AI image model that gives users finer control over editing photos, a step meant to catch up with OpenAI\u2019s popular image tools<\/a> and draw users from ChatGPT.<\/p>\n

The update, called Gemini 2.5 Flash Image, rolls out starting Tuesday to all users in the Gemini app, as well as to developers via the Gemini API, Google AI Studio, and Vertex AI platforms.<\/p>\n

Gemini\u2019s new AI image model is designed to make more precise edits to images \u2014 based on natural language requests from users \u2014 while preserving the consistency of faces, animals, and other details, something that most rival tools struggle with. For instance, ask ChatGPT or xAI\u2019s Grok to change the color of someone\u2019s shirt in a photo, and the result might include a distorted face or an altered background.<\/p>\n

\"anGemini 2.5 Flash\u2019s native image editor blends photos of a dog and person, while keeping their likeness. Credit: Google<\/p>\n

Google\u2019s new tool has already drawn attention. In recent weeks, social media users raved<\/a> over an impressive AI image editor in the crowdsourced evaluation platform, LMArena. The model appeared to users anonymously under the pseudonym \u201cnano-banana.\u201d <\/p>\n

Google says it\u2019s behind the model (if it wasn\u2019t obvious<\/a> already<\/a> from all the banana-related hints), which is really the native image capability within its flagship Gemini 2.5 Flash<\/a> AI model. Google says the image model is state-of-the-art on LMArena and other benchmarks.<\/p>\n

\"aGoogle claims its new AI image model is state-of-the-art on several benchmarks. CREDIT: GOOGLE<\/p>\n

\u201cWe\u2019re really pushing visual quality forward, as well as the model\u2019s ability to follow instructions,\u201d said Nicole Brichtova, a product lead on visual generation models at Google DeepMind, in an interview with TechCrunch. <\/p>\n

\u201cThis update does a much better job making edits more seamlessly, and the models outputs are usable for whatever you want to use them for,\u201d said Brichtova.<\/p>\n

AI image models have become a critical battle ground for Big Tech. When OpenAI launched GPT-4o\u2019s native image generator in March, it drove ChatGPT\u2019s usage<\/a> through the roof thanks to a frenzy of AI-generated Studio Ghibli<\/a> memes that, according to OpenAI CEO Sam Altman, left the company\u2019s GPUs \u201cmelting<\/a>.\u201d<\/p>\n

To keep up with OpenAI and Google, Meta announced last week that it would license<\/a> AI image models from the startup Midjourney. Meanwhile, the a16z-backed German unicorn Black Forest Labs<\/a> continues to dominate benchmarks with its FLUX AI image models.<\/p>\n

Perhaps Gemini\u2019s impressive AI image editor can help Google close its user gap with OpenAI. ChatGPT now logs more than 700 million<\/a> weekly users. On Google\u2019s earnings call in July, the tech giant\u2019s CEO Sundar Pichai revealed that Gemini had 450 million<\/a> monthly users \u2014 implying weekly users are even lower.<\/p>\n

Brichtova says Google specifically designed the image model with consumer use cases in-mind, such as helping users visualize their home and garden projects. The model also has better \u201cworld knowledge\u201d and can combine multiple references in a single prompt; for example, merging an image of a sofa, a living room photo, and a color palette into one cohesive render.<\/p>\n

\"anGemini 2.5 Flash Image lets users have \u201cmulti-turn\u201d conversations with an AI image model. CREDIT: GOOGLE<\/p>\n

While Gemini\u2019s new AI image generator makes it easier for users to make and edit realistic images, the company has safeguards that limit what users can create. Google has struggled with AI image generator safeguards in the past. At one point, the company apologized<\/a> for Gemini generating historically inaccurate pictures of people, and rolled back<\/a> the AI image generator altogether.<\/p>\n

Now, Google feels that it\u2019s struck a better balance. <\/p>\n

\u201cWe want to give users creative control so that they can get from the models what they want,\u201d said Brichtova. \u201cBut it\u2019s not like anything goes.\u201d<\/p>\n

The generative AI section of Google\u2019s terms of service prohibits users from generating \u201cnon-consensual intimate imagery.\u201d Those same kinds of safeguards don\u2019t seem to exist for Grok, which allowed users to create AI-generated explicit images<\/a> resembling celebrities, such as Taylor Swift. <\/p>\n

To address the rise of deepfake imagery, which can make it hard for users to discern what\u2019s real online, Brichtova says that Google applies visual watermarks to AI-generated images, as well as identifiers in its metadata. However, someone scrolling past an image on social media may not look for such identifiers.<\/p>\n","protected":false},"excerpt":{"rendered":"Google is upgrading its Gemini chatbot with a new AI image model that gives users finer control over…\n","protected":false},"author":3,"featured_media":177328,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[7],"tags":[100577,100578,39852,2722,158,67,132,68],"class_list":{"0":"post-177327","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-technology","8":"tag-ai-image","9":"tag-ai-image-model","10":"tag-gemini","11":"tag-google","12":"tag-technology","13":"tag-united-states","14":"tag-unitedstates","15":"tag-us"},"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@us\/115095743232304574","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/posts\/177327","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/comments?post=177327"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/posts\/177327\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/media\/177328"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/media?parent=177327"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/categories?post=177327"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/tags?post=177327"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}