Gemini app expands to audio files (www.theverge.com)

Google has expanded the capabilities of its Gemini-powered AI suite with three major updates, the most notable being the Gemini app's new support for audio file uploads—the top user request. Free users can now process up to 10 minutes of audio and five prompts daily, while paid AI Pro and Ultra subscribers can handle up to three hours per audio file, supporting multiple formats including zipped files. This expansion allows users to leverage Gemini’s multimodal understanding to analyze or query voice content seamlessly, marking a significant step toward more versatile AI interactions. In addition, Google Search’s AI Mode has broadened language support by integrating Gemini 2.5, now accommodating Hindi, Indonesian, Japanese, Korean, and Brazilian Portuguese. This enhancement enables a wider global audience to ask complex questions in their native languages and receive nuanced search results. Meanwhile, Google’s NotebookLM, an AI-powered research assistant, has introduced customizable report generation in over 80 languages. Users can create various content formats such as blog posts, study guides, flashcards, and quizzes, tailoring tone and style based on uploaded documents and media—transforming it into a flexible tool for comprehensive knowledge synthesis. Together, these updates illustrate Google’s push toward richer, multimodal AI experiences that serve a global user base with diverse content formats. By enhancing audio input in Gemini, multilingual search capabilities, and adaptive report generation in NotebookLM, Google is advancing the accessibility and utility of AI-driven workflows across research, education, and everyday information retrieval.