Google boosts Gemini 2.5 Flash: improvements in reasoning, visual understanding, and efficiency Google has launched a significant update for Gemini 2.5 Flash, its artificial intelligence model, which greatly enhances its reasoning capabilities, image comprehension, and operational efficiency. This new version is now available to all users through the Gemini app, both on the web and mobile devices. Among the most notable features, Gemini 2.5 Flash now provides clearer and more detailed explanations, especially useful for school tasks or complex topics. Thanks to its improvements in reasoning, the AI can break down concepts step by step, making it easier to understand any query. The update also introduces an optimized response format: Gemini uses headers, lists, and tables to organize information in a more visual and accessible way. Additionally, when uploading images or diagrams, the model can analyze, explain, and even convert them into study cards, thus expanding its multimodal capabilities. To access these improvements, simply select the “2.5 Flash” option in the model selector within the Gemini app. No additional configuration is required, as the new version automatically replaces the previous one. Along with the launch for end users, Google has introduced preview versions of Gemini 2.5 Flash and Gemini 2.5 Flash Lite for developers, available in Google AI Studio and Vertex AI. According to the company, the preview version of Gemini 2.5 Flash offers a 5% performance boost in complex applications, thanks to greater efficiency in the use of agentic tools, lower latency, and reduced computational resource consumption. On the other hand, Gemini 2.5 Flash Lite is optimized for following complex instructions, generating more concise responses, and further reducing latency. This version also enhances multimodal and translation capabilities, allowing for audio transcription, image interpretation, and more accurate text translation. It is important to note that these preview versions are exclusively for developers and, while not completely stable, offer greater speed and efficiency than standard models. With this update, Google reaffirms its commitment to a more powerful, versatile, and accessible AI for both users and developers.