TLDR
- Google launches enhanced Gemini AI capabilities across Docs, Sheets, Slides, and Drive for AI Ultra and Pro subscription tiers.
- Gemini in Docs creates complete documents, mirrors writing styles, and copies formatting from existing reference materials.
- Spreadsheet automation arrives in Sheets through natural language commands, including live Google Search data integration.
- Gemini Embedding 2 debuts as a multimodal system processing text, images, video, audio, and PDF files within a single framework.
- Alphabet (GOOGL) shares posted modest gains during Tuesday’s midday session.
Google continues integrating Gemini throughout its productivity ecosystem. Tuesday’s announcement revealed expanded Gemini AI functionality across Google Docs, Sheets, Slides, and Drive — launching in beta form for AI Ultra and Pro subscription holders with immediate availability.
The deployment reaches English-speaking users worldwide for Docs, Sheets, and Slides. Drive’s upgraded features currently serve U.S. markets exclusively, with additional language support scheduled for future releases.
Docs users can describe their requirements, prompting Gemini to create documents by drawing from stored files, email messages, and web resources. The system replicates writing styles from other documents and transfers formatting from reference files — enabling tasks like auto-populating travel itinerary templates using flight and accommodation information extracted from email correspondence.
Sheets receives comparable functionality. Users issue prompts directing Gemini to construct complete spreadsheets. Google demonstrated this with a relocation checklist that extracts contact information from emails and monitors price quotes arriving in your inbox.
The Fill with Gemini capability extends these functions — retrieving current data from Google Search to complete tables. Examples include college application deadlines or tuition costs, automatically populated without manual entry.
Slides and Drive Also Get Updates
Within Slides, Gemini creates individual slides matching existing presentation themes and color schemes. The system draws context from files, emails, and online sources. Google confirms that complete presentation generation via single prompts remains under development.
Drive’s Ask Gemini function adds AI-created summaries appearing above search results, extracting information from files with proper citations. Users can search across documents, email, calendar entries, and web content from a unified interface.
Gemini Embedding 2 Also Launched Tuesday
Google simultaneously introduced Gemini Embedding 2, a multimodal system processing text, images, video, audio, and documents within one unified embedding framework.
The architecture accommodates up to 8,192 input tokens for text processing, manages up to six images per request in PNG or JPEG formats, and handles up to 120 seconds of video content in MP4 or MOV formats.
Audio ingestion occurs directly — bypassing transcription requirements — while PDF embedding supports documents extending to six pages.
Google reports the model handles over 100 languages and targets applications including semantic search, sentiment analysis, and Retrieval-Augmented Generation (RAG).
The company claims performance advantages over Amazon and Voyage models across text, image, and video processing — based on Google’s internal benchmarking.
Alphabet (GOOGL) shares registered fractional gains during Tuesday’s midday trading period.
