Google today gave Gemini Notebook users a way to inspect and change the plan behind two of the product's generated formats, Slide Decks and Video Overviews, before anything is produced. A new Review toggle exposes "outline notes" for slides and "producer notes" for videos, and the co-creation flow is rolling out globally to all Gemini Notebook users over the coming week.

In Short

Gemini Notebook, the Google tool formerly called NotebookLM, can turn uploaded documents into slides or short narrated videos, and until now it went more or less straight from a request to the finished result. From today, it can first propose a list of topics and a plan, which people can change before Google's system builds the final slides or video. That matters for anyone who uses the tool to prepare decks or explainers from research, because a wrong structure can now be caught before the output exists rather than after.

What Google published

The announcement is short. Dated October 19, 2026 and labelled a one-minute read on Google's blog, it is credited to Viraj Doshi, Senior Product Manager for Gemini Notebook, and Shan Wang, a Senior UX Designer on the same product. It runs to roughly a dozen sentences, accompanied by two embedded walkthrough clips titled "Co-creation Slides walkthrough" and "Co-creation Video Overview walkthrough," whose contents are not reproduced in the text.

Google frames the change as a response to users. "Based on your feedback, Gemini Notebook now offers more creative control from start to finish on two of our popular outputs: Slide Decks and Video Overviews," according to the post. The core of the change is described in a single sentence: "Instead of jumping from an initial prompt straight to a final output, Gemini Notebook can now walk you through the creation flow in a quick process based directly on your sources."

That phrase, "based directly on your sources," is the defining trait of the product. Gemini Notebook answers from the documents a user uploads to a notebook rather than from the open web, a form of grounding that Google has used to distinguish the tool from its general-purpose Gemini chatbot since the product's early days under its previous name.

How the new flow works

The post describes three routes into the two formats. They differ in how much the user intervenes before generation.

Starting from the chat

The first route begins in the notebook's chat window. A user asks Gemini Notebook to create slides or a Video Overview, and the system, according to Google, "will suggest topics to be included in your artifact" drawn from the uploaded sources. After the user confirms those topics, the post says, it becomes possible "to steer the content further and guide visual style and format."

The post does not list the visual styles or formats on offer. It does not say how many topics the system proposes, whether suggested topics can be deleted as well as added, or whether the topic list reflects every source in a notebook or only some of them.

The Review toggle

The second route goes further. Google calls it "our new Review toggle," and it is the only control the post names as new. With the toggle on, the plan for the output appears in the studio panel, the area of the interface where Gemini Notebook's generated formats are created and stored. Slides carry a button labelled "Review outline notes." Videos carry one labelled "Review producer notes."

At that stage, according to Google, users can "give additional feedback, like asking to include an executive summary at the beginning or swapping the order of a scene, before anything renders." A button labelled "Proceed" then produces the final output.

The two examples are telling. An executive summary at the front of a deck and a reordered scene in a video are structural edits, the kind that previously would have required generating the whole output again with a revised prompt and hoping the new version came out in the intended order. The post does not say whether the outline and producer notes can be edited directly as text, or whether all changes pass through further instructions to the model.

Skipping straight to review, or skipping review entirely

The third and fourth variants are shortcuts at either end. According to the post, a user who "already know[s] what you want to create" can say so in the prompt, and "the notebook will go straight to the review flow," bypassing the topic suggestion step.

At the other extreme, the old path remains. Users can still open the studio panel, click the output type, leave the Review toggle off and "have Notebook generate the artifacts, as usual," according to Google. Nothing in the post suggests the review step will become mandatory.

Availability

The rollout statement is one sentence: "These co-creation capabilities are rolling out globally to all Gemini Notebook users over the upcoming week."

Two words in it carry weight. "Globally" implies no country restriction, and "all" implies no tier restriction. That stands in contrast to the most recent large Gemini Notebook update PPC Land covered. When NotebookLM moved to Gemini 3.5 and gained a cloud computer and 11 downloadable output formats on June 8, access was limited at launch to Google AI Ultra subscribers and Workspace customers with AI Ultra Access or AI Expanded Access, with free-tier users excluded.

What the sentence leaves out is just as relevant. The post does not mention:

  • whether the flow works in the Gemini Notebook mobile apps or only on the web;
  • which languages the topic suggestions and outline notes support;
  • whether Workspace business and education accounts receive it on the same schedule;
  • whether a reviewed plan that is later abandoned, or a plan revised several times, counts against usage limits.

The last point is not academic. Among the related stories Google's blog listed beneath the post was one titled "We're introducing flexible usage limits for Gemini Notebook." Its contents are not part of the attached material and are not covered here. Elsewhere in Google's consumer AI products, limits have become more granular: PPC Land reported that from October 9 the Gemini app restricted free accounts to Flash-Lite, with compute-based limits that refresh every five hours until a weekly cap is reached, and multipliers of 2x and 4x standard usage for the AI Plus and AI Pro plans.

How the two formats got here

Video Overviews have a documented history on PPC Land; Slide Decks less so. Google's post calls both "popular outputs" but does not say when Slide Decks were introduced or how widely either is used.

Google launched Video Overviews on July 29, 2025 as narrated slides combining generated visuals with spoken explanation, initially in English only. Customisation at that point consisted of prompt instructions about focus topics, learning objectives, audiences and expertise levels, the same levers available for Audio Overviews. Within a month, Video Overviews expanded to more than 80 languages by August 25, 2025.

The next change concerned appearance rather than structure. On October 13, 2025, Google brought its Nano Banana image model into Video Overviews, adding six visual styles such as watercolor and anime, and a condensed "Brief" format alongside the standard one. Users picked from predefined styles rather than writing their own.

Then came the rename. Google renamed NotebookLM to Gemini Notebook on July 16, 2026, citing more than 30 million individual users and more than 600,000 organisations, while keeping it as a standalone product and folding it into the Gemini app and Google Search.

Read against that sequence, today's update fills a gap. Earlier changes expanded where Video Overviews worked and how they looked. Control over what the output contained, and in what order, still ran through a single prompt and a single generation. The Review toggle adds an intermediate, inspectable plan between the two.

Why the review step matters

Generated presentations and videos fail in particular ways. A large language model summarising a set of documents may weight a minor source too heavily, skip a section a user regards as central, or order the material in a way that makes sense statistically but not rhetorically. Outright errors are possible as well, and hallucination in a slide is harder to spot than in a chat reply, because it is wrapped in finished design and narration.

A reviewable outline does not remove those risks, but it moves the point of inspection earlier. A topic list or a set of producer notes is shorter and plainer than a finished deck, so mistakes of selection and sequence are easier to see. It is also, presumably, cheaper for Google to regenerate a plan than a full video, although the post makes no claim about compute and that inference is PPC Land's, not Google's.

The language in the interface is notable too. "Producer notes" borrows from video production, where a producer's brief fixes the running order before anything is shot. "Outline notes" is the presentation equivalent. Google is presenting the model as an assistant that drafts a brief for approval rather than a generator that returns a finished product unprompted.

What it means for marketing teams

The relevance for advertising and marketing work is practical rather than strategic. Agencies, in-house teams and consultants already use document-grounded tools to turn research, briefs, transcripts and reports into first-draft decks and internal explainers. A structure that matches a client's expectations, an executive summary up front for example, is often the difference between a usable draft and one that has to be rebuilt.

The update also lands in a busy month for Google's AI products. PPC Land reported on October 13 that Gemini 4 Argon reached the Gemini app for AI Pro and Ultra subscribers along with a task mode that keeps working while users are away, and that Google priced Gemini 4 Argon at $2 per million tokens for paid users only. Today's post does not say which model generates the topic suggestions, outlines or final outputs in Gemini Notebook, and it does not mention Argon at all.

Some questions remain open until the rollout completes and Google documents the feature in its help pages. Whether the outlines can be exported, whether a reviewed plan can be reused across notebooks, whether teams on shared notebooks can review a plan together, and how the reviewed outline relates to the PowerPoint export added in June are all absent from the post. So is any measure of how often users previously had to regenerate outputs, the problem the feature appears designed to address.

Timeline

Summary

Who: Google's Gemini Notebook team, with the post credited to Senior Product Manager Viraj Doshi and Senior UX Designer Shan Wang. The change affects all Gemini Notebook users, including the marketing and agency teams that use the tool to turn research into decks and videos.

What: A co-creation flow for Slide Decks and Video Overviews. Gemini Notebook suggests topics from a notebook's sources, and a new Review toggle shows "outline notes" for slides and "producer notes" for videos in the studio panel, where changes such as adding an executive summary or reordering scenes can be requested before a "Proceed" button generates the final output. The previous one-click path remains available.

When: Google published the post today, October 19, 2026, with a global rollout to all users over the following week.

Where: In Gemini Notebook, through its chat window and studio panel. The post does not say whether the mobile apps are included.

Why: Google says the change responds to user feedback asking for more creative control. It places a reviewable plan between the prompt and the finished output, so problems of structure and selection can be corrected before generation rather than by regenerating the whole artifact.