Google’s NotebookLM Released 5 New Features That Are Insane (Create Slides, Infographics, Deep Research and More)

NotebookLM just unlocked a set of features that dramatically expand what you can do with your research and content. If you work with notes, presentations, marketing assets, or AI workflows, these updates will speed up production and let you produce polished outputs without leaving NotebookLM. Below I explain each new capability, share practical workflows and prompt tips, and show how to get predictable, professional results fast.

Table of Contents

• What’s new at a glance
• Why these updates matter
• Feature deep dive
• Practical workflows that use these features together
• Step-by-step: Create a slide deck in NotebookLM (quick guide)
• Prompt examples that work well
• Quality checks and export advice
• Suggested images and multimedia to include
• How do I export slides and infographics for editing?
• Can I control which sources NotebookLM uses to generate content?
• What’s the difference between fast research and deep research?
• How do conversation goals affect NotebookLM’s responses?
• Is the infographic text proofed for spelling and grammar?
• Meta description
• Tags and categories to use
• Final notes and call to action

What’s new at a glance

• Built-in slide creation: Generate either detailed slide decks or clean presenter slides from your sources.
• Infographic generator: Create shareable, social-ready infographics with orientation and style options.
• Video overview customization: Auto-create short explainers or brief visual summaries with selectable visual themes.
• Deep research mode: Import many more sources at once and produce a long-form research report.
• Chat and conversation improvements: Refresh chats, set conversation goals and preferred response length, and keep chat histories saved for later editing.

Why these updates matter

These features move NotebookLM from an exploratory note tool into a content-creation and collaboration platform. Instead of copy-pasting research into PowerPoint, a design tool, or a script editor, you can now generate polished deliverables directly from your curated sources. That reduces friction, speeds iteration, and makes it easier to reuse the same research across multiple formats—slides for a meeting, an infographic for social, and a short video overview for a quick explainer.

Feature deep dive

1. Slide creation: detailed decks and presenter slides

NotebookLM now generates two types of slide outputs:

• Detailed slide presentations: These are text-rich decks suitable for distribution—think handouts, classroom notes, or in-depth internal reports.
• Presenter slides: Lightweight, visual slides with concise bullet points and clear visuals to support live speaking or recording.

Key controls you should use:

• Length setting – choose short or long to control verbosity.
• Detail level – pick default or custom styles, and specify the exact format (e.g., Google Slides, PowerPoint export-ready).
• Select sources – include all sources or pick a subset so you can produce several different slide decks on the same topic by changing which sources are used.

Practical tips:

• For a meeting presentation, use presenter slides with short bullets and speaker notes. For a distribution deck, use the detailed option with more context on each slide.
• Specify style prompts such as color palette, font weight, or brand tone in the description box before you generate. Example prompt: “Create 8 slides in a clean blue-and-white brand palette with a modern sans-serif tone and speaker notes for each slide.”
• If a slide generation contains a minor error, regenerate the deck or adjust the source selection. You can also export to Google Slides or PowerPoint for manual tweaks.

2. Infographics: social-ready visual summaries

The infographic generator is a big deal for content creators and marketers. It turns your research into a visual asset suitable for LinkedIn, Instagram, TikTok thumbnails, or blog hero images.

What you can control:

• Orientation: Landscape, portrait, or square depending on the platform.
• Level of detail: Concise (fast), standard, or detailed (slower but more nuanced).
• Style and color: Give guidance on color schemes, iconography, and overall style (minimal, retro, corporate, playful, etc.).
• Source selection: As with slides, choose which sources to base the infographic on.

Use cases and examples:

• Create a concise portrait infographic for Instagram that explains a concept in four panels.
• Generate a landscape infographic for LinkedIn summarizing a research report with statistics and a short conclusion.
• Transform a technical breakdown into a shareable “anatomy” infographic—an example is an AI agent breakdown with Perception (trigger), Planning (brain), Action (hands), and Memory (output).

Pro tips:

• Choose concise when you need speed and standard or detailed when you want richer captions or more design elements.
• Ask for consistent iconography and a limited palette to ensure visual coherence across multiple infographics.
• Download and run a quick spell check after generation—infographic text is usually clean, but it is good to confirm brand terms and acronyms are accurate.

3. Video overview customization: explainer or brief with style control

NotebookLM can auto-generate a short video overview or explainer for any topic you feed it. You choose whether it should be brief (bite-sized) or an explainer (structured and more detailed). You also pick a visual theme to control the style of the generated video.

How to use it:

• Pick brief for quick summaries or social shorts; pick explainer for structured multi-minute explainers suitable for training or onboarding.
• Select a visual style—retro 90s, futuristic, corporate, or custom—and the tool will apply that look to animations and visuals.
• Adjust the script by editing the text prompt before generating. Small prompt edits can tighten pacing or change emphasis.

Practical example workflow:

1. Use deep research to gather sources about a topic (see next section).
2. Create a brief video overview with a chosen theme and ask for a 2–3 minute length.
3. Review the generated narration and visuals. Trim pauses by editing the script or scene timing before exporting.

Tip: A two-to-three minute overview is ideal for repurposing as a podcast episode, social clip, or an intro for a course module.

4. Deep research mode: more sources, stronger reports

NotebookLM now offers two research modes when you discover sources: fast research (quick, ~7–10 sources) and deep research (comprehensive, dozens of sources). Deep research takes longer but returns far more citations and produces a long-form research report you can import into your notebook.

Why use deep research:

• Get a broader evidence base for decision-making or academic work.
• Produce long-form reports with full citations and structured sections that can be exported or used to generate slides, infographics, or videos.
• Capture a wider range of viewpoints and reduce bias from a small set of sources.

How to combine deep research with content generation:

1. Use deep research to populate a notebook with 30+ sources on your topic.
2. Select the relevant subset of sources when generating slides or infographics to focus the narrative.
3. Create a video overview and tell NotebookLM the target audience and visual style for consistency across assets.

Example: Searching “How the TikTok algorithm works” on fast research might return seven sources. Deep research can return 30+ sources and a full report with citations ready to build a classroom lesson, a slide deck, and a short explainer video from the same research base.

5. Chat improvements: refresh, conversation goals, saved chats

Several small-seeming changes here have a big impact on workflow stability and collaboration:

• Refresh chat clears chat history quickly when you want a fresh conversation or you want to remove sensitive content before sharing.
• Conversation goals let you set the intent of the interaction—study for school, prepare for a meeting, generate marketing copy, etc. NotebookLM then tailors responses accordingly.
• Response length control lets you choose short, medium, or long answers so the system responds in the format you prefer.
• Saved chat persistence means the chat remains accessible even if you navigate away—no more accidental loss from browser backspace or reload.

Why this matters:

Being able to define the goal and response length produces outputs that are immediately usable. No more manual edits to convert a long conversation into a meeting-ready agenda or a study-friendly summary. Refreshing the chat before sharing also protects sensitive lines of inquiry or private notes.

Practical workflows that use these features together

Content repurposing for a single research project

1. Run deep research on your topic and import the generated report into a notebook.
2. Generate a detailed slide deck for distribution to stakeholders.
3. Generate a concise infographic (square or portrait) for social distribution.
4. Make a two-minute video overview with a consistent visual theme to use as a promo or training intro.

This single-research approach creates a consistent, branded set of assets ready to publish across channels.

Teaching and training

• Create a detailed slide deck for students with supplementary speaker notes and references.
• Produce a short explainer video as pre-class material.
• Share the notebook with co-instructors and use analytics to see who engaged with the content.

Team knowledge sharing

• Use the conversation goal feature to craft meeting prep across team roles (design, marketing, product).
• Generate presenter slides for the meeting lead and a detailed deck for attendees.
• Refresh chats with sensitive items before sharing the notebook externally.

Step-by-step: Create a slide deck in NotebookLM (quick guide)

1. Open your notebook and confirm your sources. Use deep research if you need many sources.
2. Click Slide and choose Detailed or Presenter depending on your use case.
3. Choose short or long length, and provide a style prompt that includes colors and format (e.g., “8 slides, corporate blue palette, include speaker notes”).
4. Select the specific sources to include so the deck focuses on the right evidence.
5. Click Generate. Review slides, regenerate a slide if something’s off, or export to Google Slides or PowerPoint for final edits.

Prompt examples that work well

• Slide deck prompt: “Create 10 presenter slides summarizing the product roadmap. Use concise bullets, brand colors: #003366 and #00aaff, include speaker notes (1 sentence per slide).”
• Infographic prompt: “Portrait infographic, concise mode. Use four panels to explain the AI agent anatomy: Perception, Planning, Action, Memory. Minimal icon style, color palette coral and charcoal.”
• Video overview prompt: “Brief explainer, 2–3 minutes, retro 90s theme. Focus on how to build a basic AI agent with steps and examples.”

Quality checks and export advice

• Always proof the generated text for brand names, acronyms, and licensing requirements.
• For infographics and slides, export and open in the target tool for final alignment and fonts.
• Use source selection strategically to produce multiple perspectives or different lengths of content without redoing research.

Suggested images and multimedia to include

• Hero image: screenshot of a generated infographic with descriptive alt text such as “Infographic showing AI agent components: Perception, Planning, Action, Memory.”
• Slide preview: show a sample deck slide with alt text like “Presenter slide example with bullet points and speaker notes.”
• Short video clip: brief example of a generated video overview to increase engagement.

Meta description

Create slides, infographics, and deep research reports with NotebookLM’s five new features. Learn workflows, prompt examples, and pro tips to produce polished assets fast.

Tags and categories to use

Tags: NotebookLM, Google NotebookLM, AI tools, AI productivity, slides, infographics, deep research, video overview, content automation

Final notes and call to action

NotebookLM is rapidly evolving into a full content production assistant—research, visuals, and short-form video from a single workspace. Use deep research to build a rich foundation, then repurpose that foundation across slides, infographics, and videos. Be deliberate about source selection, style prompts, and response length to get outputs that require minimal manual editing.

If you want to automate more of your work with AI, pick one small project and run it through this end-to-end workflow: research, generate a detailed deck, create a social infographic, and make a short video overview. Iterate until your prompts deliver consistent results, and then scale the process across other topics.