Google claims to have created a viral image processing program that caused huge waves on the Internet. And it's a bad day for Adobe. Google announced on Tuesday that Google Deepmind developed the model and integrated it into the Gemini app and website, making it available to all users. This model, originally codenamed "Nano Banana" for testing, is now officially named Gemini 2.5 Flash.
Google has tested the new model on the LMArena platform. This public website is dedicated to collecting anonymous feedback on the quality of AI models. Although the test was named "Nano Banana", users quickly speculated that the developer behind it was Google.
According to Google, the core advantage of Gemini 2.5 Flash is that it can maintain the consistency of characters or animals in different images. Based on existing feedback data, based on the Elo scoring system that measures the relative performance of generative models, Google claims that users’ preference for the new model has surpassed OpenAI and other competitors.
This could pose a significant threat to existing image processing programs, especially products such as Adobe's Photoshop.
Adobe responded urgently on Tuesday, announcing that it would immediately integrate Google's new model in Adobe Firefly and Adobe Express. In response to the question "Why do users choose Adobe tools instead of using Google directly?" Adobe emphasized that its advantage lies in providing one-stop model integration services.
However, this move has not been able to effectively alleviate investors' concerns so far. Adobe's stock price fell 35% last year, partly due to analysts' concerns about the impact of AI on the traditional tool market.