Google announced on August 5 that it has completed the rollout of a large-scale update for the Pro plan of its AI research tool “Gemini Notebook” (formerly NotebookLM) to all users. In addition to enhanced chat functionality, new output features have been added that can automatically generate charts, PDFs, spreadsheets, images, and more based on source materials, allowing users to complete the entire workflow from information organization to deliverable creation within a single environment. The Pro plan is included in “Google AI Pro,” priced at ¥2,900 per month (approximately $18).
At the time Google announced this update on July 16, the company disclosed that Gemini Notebook was already being used by over 30 million users and more than 600,000 organizations (as reported by U.S. media outlet 9to5Google). The service name has changed several times since its announcement at Google I/O 2023, with the evolution as follows:
Gemini Notebook is a research tool that answers questions based solely on materials uploaded by the user. In this enhanced version, each notebook is equipped with a dedicated execution environment called a “secure cloud computer,” evolving into a system where the AI can write and execute its own code. This represents a significant functional leap from simply summarizing content to calculating figures within documents and deriving analytical results.
The new output features support a wide range of use cases, from educational applications—such as organizing lecture notes into a PDF study guide with five practice questions—to business applications like calculating sales growth rates, generating charts, and drafting messaging proposals based on competitive research. Generated deliverables can be downloaded directly from the Studio panel, and users can request revisions through chat to regenerate them.
FeaturePrevious (NotebookLM)New (Gemini Notebook Pro)Brand NameNotebookLM (June 2023–)Gemini Notebook (renamed July 16, 2026)Execution EnvironmentChat responses onlySecure cloud computer with automatic code executionOutput FormatsPrimarily summaries and chat answersPDF, spreadsheets, images, interactive quiz appsMaterial CollectionUser-uploaded materials onlyCan start with vague queries and auto-collect via Google SearchUsage Scale—30M+ users, 600K+ organizations (as of July 2026, per Google)
Notably, the quiz generation feature is implemented in a distinctive way. When practice questions are requested, the tool provides not just a list of questions but an interactive quiz app assembled with specified difficulty levels and question counts. According to verification by IT news site Impress, Gemini Notebook’s thought process included records of comparing whether to output a printable PDF or an interactive format that can be solved on the spot, with the final output format being determined by the AI itself rather than the user’s instructions.
Convenience has also improved on the material preparation front. Even when users don’t have materials readily available, starting with a vague query allows Gemini Notebook to search for relevant materials via Google Search and automatically add them as information sources.
The technical foundation supporting this execution environment employs Google’s latest AI model “Gemini 3.5” and the agent-based development platform “Antigravity.” Additionally, over 100 software skills (equivalent to standard operating procedure documents) are built in. When asked about the programming languages it uses, Gemini Notebook itself responds that it uses Python 3.12 and Bash, though this is not documented in official help materials and should be treated as the AI’s self-reported information for now.
Google’s official X account has encouraged even free-tier users to bookmark posts demonstrating these new features, hinting at a future expansion to the free plan. However, specific timing and scope of availability have not been disclosed.
Meanwhile, there are noteworthy developments in Google’s hardware strategy. According to U.S. outlet TechSpot, press images have leaked of a Lenovo-manufactured “Googlebook” laptop running “Aluminium OS,” a new operating system that fuses Android and ChromeOS, which Google announced in May of this year. The Lenovo “Googlebook 15” is expected to launch in the second half of 2026, featuring a Google “G” key on the lower left of the keyboard designed to instantly launch the Gemini chatbot.
According to U.S. outlet Android Authority, the first wave of Googlebook lineup extends beyond Lenovo, with major PC manufacturers including Acer, ASUS, Dell, and HP all reportedly developing models. The Lenovo “Googlebook 15” is widely expected to be officially unveiled at IFA 2026, the consumer electronics trade show held in Berlin in September.
This G key follows a similar concept to Microsoft’s “Copilot key,” but the Googlebook strategy aims for deeper AI integration. Google positions the Googlebook as “the most advanced laptop,” emphasizing “intelligence” as the most critical performance metric over traditional specs like processing power and memory.
PCWorld points out that this high degree of automation and reliance on agent-based AI could make the Googlebook the first wave of the “anti-personal computer”—a view suggesting that devices will operate autonomously based on user intent, marking the end of an era where users independently control their machines.
The enhancement of Gemini Notebook and the emergence of the Googlebook demonstrate that Google is accelerating the reconstruction of an AI-centric ecosystem spanning from personal learning and work environments to hardware platforms. The vision of cloud-based AI agents autonomously executing everything from document analysis to deliverable creation, with results seamlessly utilized on dedicated devices, is rapidly becoming a reality.
In educational settings, the ability to automatically generate quizzes at varying difficulty levels from lecture materials could significantly reduce the workload of teachers in creating instructional materials. In the business realm, the tool also supports advanced analytical tasks such as creating real estate analysis tables and automatically generating competitive research reports, holding the potential to fundamentally transform how knowledge workers operate.