You're not generating slides. You're generating arguments. Slides are how the argument ships.
NotebookLM can produce a deck in 90 seconds. That's the easy part. The hard part is knowing what the deck has to do — who decides what after seeing it, what objection it has to neutralize, what action it has to extract. This page is the strategic framework that goes before the tactics. Make the 4 decisions first, then pick the workflow.
★ The Boardroom Brief Generator — Featured Free Prompt
You are the strategic editor preparing a deck brief BEFORE any slides get generated. I will not let you produce slide content yet. Your job is the brief.
Inputs you have: the sources in this notebook. Inputs you need to extract from me: nothing. You will infer audience and intent from the source material plus the explicit task statement below.
EXPLICIT TASK: [Describe the deck's purpose in one sentence — what decision, action, or alignment you need to drive.]
Return a 7-section strategic brief:
SECTION 1 — THE DECISION BEING ASKED FOR
State, in one sentence, the single decision or action the audience must take after seeing this deck. Not "be informed." Not "understand." A concrete decision: approve the budget, sign the partnership, adopt the framework, redirect the team.
SECTION 2 — THE AUDIENCE TENSION
What is the audience most likely to push back on, and why? Name the specific objection they will raise in the first 60 seconds.
SECTION 3 — THE EVIDENCE ARSENAL
From the sources in this notebook, identify the 3 strongest pieces of evidence for our position. For each: (a) one-sentence summary, (b) which source and which page, (c) what objection it neutralizes.
SECTION 4 — THE 7-SLIDE SPINE
Propose the 7-slide structure that walks from audience tension to the decision being asked. For each slide: (a) the headline (action-oriented sentence, not a topic), (b) the single argument the slide makes, (c) the evidence it cites.
SECTION 5 — THE OPENING MOVE
Propose the opening move: a number that surprises, a one-sentence reframe of the problem, or a question that traps them into the conversation we want.
SECTION 6 — THE CLOSING CLOSE
Write the exact closing sentence the presenter will say — the line that turns a deck into a decision. One sentence. No hedge words.
SECTION 7 — PRE-MORTEM
Name the single most likely failure mode and propose the one specific change that prevents it.
End with one line: "Generate the deck now" — or, if the brief reveals the deck does not yet have a coherent argument, "Stop. Rework the source material before generating."
Free · No credit card · Run on any NotebookLM notebook with research sources
4
Strategic decisions — audience, action, format, iteration mode
90s
Generation time — basic decks in 30-60s, complex up to 10 min
35
FAQs answered — every question about NotebookLM slides
v2 · Updated June 2026
This is the consolidated master guide — it merges and expands our earlier slide-deck guides (the strategy framework and the tactical workflows) into one operating system. Everything from the previous versions lives here, now with the full PPAE pipeline and 80+ prompts.
The PPAE pipeline: 4 phases that power every workflow on this page
Every workflow on this page follows the same 4-phase PPAE architecture: Perceive (ingest and clean sources), Plan (detect gaps and build structure), Act (generate, debate, revise), Evaluate (score quality, audit bias, track progress). The agents adapt to the task — the framework is universal.
How to create slides in NotebookLM: Upload your sources (PDFs, Google Docs, YouTube URLs) to a notebook → Open the Studio panel → Click Slide Deck → Paste a custom prompt specifying audience, slide count, and tone → Click Generate. A full deck renders in 30–90 seconds for basic notebooks, up to 10 minutes for complex ones. Edit individual slides with the Pencil UI (natural-language revisions, 30 seconds per slide). Export as PPTX (editable in PowerPoint/Google Slides) or PDF. Every claim in the deck traces to your uploaded sources via retrieval-augmented generation (RAG).
Who needs a strategic deck framework before tactics?
Anyone whose deck has to drive a decision — not just inform. This page is the layer above tactics. Make the 4 decisions first, then pick the workflow.
Founder
Series B raise, 14-slide pitch
The pitch is not a company tour. It is a single argument: this round at this valuation makes sense because of this specific moment.
Executive
Board update or strategic recommendation
The board doesn't want a status report. They want a recommendation with the work shown.
Consultant
Client deliverable, McKinsey-style
The deck is the deliverable. The argument structure is the value.
Educator
Course module, 8-12 slides per topic
Different mode entirely — the "decision" is whether the learner will engage with the next module.
Why strategy precedes tactics in NotebookLM slide work
NotebookLM has gotten so good at slide generation that the bottleneck has moved upstream. A deck generated without a brief is a tour of the sources. A deck generated from a brief is a weapon for a specific room.
The full 7-section brief template plus 5 audience-specific variants. Permanent PDF, instant delivery.
No spam · Unsubscribe anytime · 14,000+ operators on the list
From source notebook to shipped deck: the strategic flow
Same notebook, 4 strategic paths, 4 specialist workflows.
How long does NotebookLM take to generate slides?
Deck Type
Sources
Time
Notes
Basic
1-3 sources
30-60 sec
No custom prompt, default formatting
Standard
3-8 sources
60-90 sec
Custom prompt, structured output
Complex
10+ large sources
3-10 min
Technical content, many citations
Heavy
40+ pages, 100+ sources
15-30 min
Runs in background, progressive render
Pro tipThe process runs in background. You can keep working in the same notebook while it generates. Slides render progressively.
Workflow 2: Instant Slide Generation — 90 Seconds
PHASE 3: ACT
NotebookLM generates a complete, source-cited slide deck from your uploaded documents in 60–90 seconds. Every claim traces to your sources. Here's the exact 6-step workflow.
01
Upload and organize your sources
Create a focused notebook. Add PDFs, Google Docs, YouTube URLs (auto-transcribed), pasted text. One topic per notebook. Pin your most important source for heavier weighting.
02
Open Studio panel → select Slide Deck
Click the Slide Deck tile. Choose: Detailed Deck (comprehensive, self-contained) or Presenter Slides (visual, talking points).
03
Paste a structured prompt
Use the free prompt below. Fill in bracketed placeholders: audience, slide count, format, density, tone. Never generate without a prompt.
04
Generate and review
Click Generate. Deck renders in 60–90 seconds. Click any underlined phrase to verify the source passage.
05
Revise specific slides
Click the pencil icon. Type natural-language revision. Queue across slides. Batch-apply in one pass. See Workflow 4: Pencil Revisions.
06
Export as PPTX or PDF
Click export icon. PPTX = editable text boxes in PowerPoint/Google Slides. PDF = fixed layout. Apply brand theme after export.
Scenario
Time
Notes
Basic (1–3 sources)
30–60 sec
Default format, minimal config
Standard (3–8 sources, custom prompt)
60–90 sec
Most common use case
Complex (10+ sources)
2–3 min
Progressive rendering
Heavy (100+ sources)
15–30 min
Split large PDFs first
Revision pass
30–60 sec
Faster than full regen
Create a [Detailed Deck / Presenter Slides] with exactly [NUMBER] slides for [AUDIENCE TYPE]. Rules: (1) Every slide title must be an action statement that states the key takeaway, not a topic label. (2) Max [3-5] bullets per slide; each must contain one specific data point from the sources. (3) After any slide with 3+ statistics, add [SUGGEST CHART: bar/line/scatter]. (4) Final slide: one bold recommendation the audience can act on tomorrow. (5) No generic stock photo suggestions. No filler slides. No slides with fewer than 2 data points. Tone: [professional/conversational/academic].
Workflow 3: The 5-Dimension Prompt Formula
PHASE 3: ACT — ADVANCED
The difference between a generic AI deck and a boardroom-ready one is not the sources — it is the prompt. Production-grade prompts specify exactly 5 dimensions. Miss any one and the output reverts to generic.
D1
Audience + Goal
"12-slide deck for a board of directors who need to approve a $2M budget increase" produces radically different output than "12 slides for a team standup."
"Max 3 bullets per slide, each under 12 words" vs. "comprehensive paragraphs with full citations." Density determines Detailed vs. Presenter.
D4
Visual Logic
"After any slide with 3+ statistics, add [SUGGEST CHART: bar/line/scatter]." Without this, the model defaults to text-only.
D5
Exclusion Rules
"No generic stock photo suggestions. No filler slides. No acknowledgment sections. No slides with fewer than 2 data points."
Style
Structure
Title Rule
Best For
McKinsey Pyramid
Answer first → evidence
Conclusion headline
Board, exec, consulting
Minimalist Pitch
Problem → solution → ask
One bold claim
Fundraising, sales
Academic Conference
Background → methods → results
Finding + citation
Conferences, defense
Education/Training
Objective → concept → practice
Learning outcome
Lectures, workshops
Design in McKinsey consulting style. Rules: (1) Every slide title must be an action statement — "Customer retention drops 23% after onboarding" not "Customer Retention." (2) Max 3 bullets per slide; each must contain one specific data point from the sources. (3) After any slide with 3+ statistics, add [SUGGEST CHART: bar/line/scatter]. (4) Final slide: one bold recommendation phrased as a single sentence. Audience: senior leadership. Tone: authoritative. Slide count: 12.
Workflow 4: Pencil UI Revisions — Fix Any Slide in 30 Seconds
PHASE 3: ACT — REVISION
You generated a 12-slide deck, 3 slides are wrong. Instead of regenerating all 12, fix only the 3 — in natural language, without losing the 9 good slides.
Dimension
Old Workflow (Pre-2026)
Pencil UI (2026)
Editing model
Element-level: click text box
Prompt-based: describe changes
Scope
One element at a time
Queue across slides, batch-apply
To fix 3 slides
Regenerate all 12 (~15 min)
Revise only 3 (~90 sec)
Risk to good slides
High — regeneration changes all
Zero — unchanged preserved
01
Generate your initial deck
Open Studio → Slide Deck → Generate. Review once without editing. Note which slides miss the mark.
02
Open the pencil on the first slide
Click pencil icon. Type revision instruction in plain language. Be directive: "Rewrite as a bold claim backed by the statistic" beats "make this more engaging."
03
Queue revisions across slides
Navigate to other slides. Add more requests. Each slide can hold multiple instructions — they stack.
04
Review the queue
Check revision queue panel for conflicts. All pending requests visible before generating.
05
Generate Revised Deck
All queued changes apply in one pass. Unchanged slides preserved exactly. Most decks reach final quality in 1 generation + 3–5 revisions.
The 6 revision types
Tone — rewrite for different audience. Structure — change layout (columns, problem/solution). Density — compress or expand. Data — add/replace specific statistics. Headline — topic label → action statement. Global — same rule across all slides.
Rewrite this slide for a senior executive audience. Lead with the business implication, not the methodology. Replace any passive constructions with direct, declarative sentences. Keep to 4 bullet points maximum — each under 12 words. Do not change any other slides. Do not add claims not present in the uploaded sources.
The closed-loop pipeline for complex research decks. 3 agents debate every slide. Confidence scoring. Bias audit. For when single-agent generation isn't enough.
Agent A
The Advocate
Generates the strongest evidence-based case for each claim. Cites best available sources. Optimizes for persuasiveness.
Agent B
The Skeptic
Attacks each claim for logical flaws. Identifies missing evidence. Flags potential biases. Rates confidence 0-1.
Agent C
The Synthesizer
Integrates Advocate and Skeptic outputs. Produces final balanced version. Assigns confidence scores. Ensures citation discipline.
Slide 5 — Advocate: "Multi-agent architectures achieve 34% higher task completion on complex software engineering tasks." [Source: Survey-2025, conf: HIGH] Slide 5 — Skeptic: "The 34% figure comes from a single survey with N=47. The confidence interval is wide. Also, Source [19] contradicts this with a 12% figure." Slide 5 — Synthesizer: "Multi-agent architectures show 20-34% improvement on complex tasks, with the range reflecting methodology differences across 3 studies. Confidence: 0.72. Caveat: limited to code generation tasks."
Phase 4: Evaluate — 5 minutes
Bias audit: mild recency bias (80% sources < 2 years). Deck-wide confidence: 0.74. 1 slide flagged (conf 0.58) for human review. Total time: 35 minutes for 50-paper, 12-slide deck.
Input: Quarterly board update for Fortune 500 client
12-slide Presenter Slides + 22-slide Detailed Deck appendix. Every claim cited by source ID. Confidence scorecard: 9/12 slides above 0.7. 3 slides revised after Skeptic flagged weak evidence. Bias audit: confirmation bias MEDIUM (only highlighted positive metrics). Fix: added risk slide with negative trends.
Final Score: 85/100 ✓
You are running a 3-agent slide debate pipeline.
AGENT A (Advocate): Make the strongest evidence-based case for [CLAIM]. Use only uploaded sources. Cite by source ID.
AGENT B (Skeptic): Challenge this claim. Identify weakest evidence link, logical leaps, selection bias. Rate confidence 0-1.
AGENT C (Synthesizer): Integrate Advocate and Skeptic outputs. Produce: (1) balanced claim, (2) confidence score 0-1, (3) caveats, (4) numbered citations.
Run this for each slide in my [N]-slide deck. Flag any slide with confidence below 0.6 for human review.
Workflow 6: Professional Pipeline — 5, 10, or 20 Minutes
Working professionals don't have 35 minutes. Choose your time mode:
5 Minutes
Emergency Deck
Meeting in 30 minutes. Upload sources → Emergency Brief → Generate → Quick scan → Export. Quality: good enough for internal review.
I have a meeting in [TIME]. I need a [N]-slide deck on [TOPIC] for [AUDIENCE]. My sources are uploaded. Generate:
1. One-sentence decision the audience must make
2. The 3 most important evidence points from my sources
3. The single biggest objection they'll raise
4. A complete [N]-slide deck with action-title headlines
Speed is priority. Use the strongest evidence available. Don't hedge — be direct.
4 niche pipelines for specialized decks
Pipeline 1
YouTube → Curriculum Deck
Add YouTube URLs (auto-transcribed), generate curriculum slides from lecture series.
Pipeline 2
VSL Production
Generate a slide deck structured as a VSL script. Screen-record with voiceover.
Pipeline 3
Conference Talk Synthesis
Combine multiple conference transcripts into a synthesis presentation.
Pipeline 4
Training Module Factory
Turn SOPs and procedures into interactive training decks with consistent format.
35 questions covering every aspect of NotebookLM slide decks.
Getting Started
Upload your sources to a notebook. Open the Studio panel, click Slide Deck, paste a custom prompt, and click Generate. A full deck renders in 30–90 seconds. Every claim traces to your sources via RAG.
Yes. Available to all users over 18 at no cost. Free tier supports up to 50 sources. NotebookLM Plus ($19.99/month) increases to 300 sources.
No. Content is generated exclusively from your uploaded sources using RAG. Every claim traces to a specific passage in your documents.
Generation
30 seconds to 10+ minutes. Basic (1-3 sources): 30-60s. Standard (3-8 sources): 60-90s. Complex (10+ sources): 3-10 min. Heavy (100+ sources): 15-30 min. Runs in background.
8-15 slides by default. You can request specific counts. AI Ultra: up to 25 slides per pass.
Detailed Deck: higher word count, stands alone, for async consumption. Presenter Slides: sparse text, for live delivery. Same notebook can generate both.
Yes. PPTX export includes speaker notes with talking points and source citations for each slide.
NotebookLM generates text-based slides. Exported PPTX has editable placeholders for adding images in PowerPoint or Google Slides.
Prompts
Specify five dimensions: audience, slide count, format, density, and tone. Include "open each slide with an action headline" and "close with a one-sentence takeaway."
80-150 words covering all 5 dimensions. Shorter = generic. Longer (200+) = model ignores later instructions.
Yes. A well-engineered prompt is a reusable template. Save it and apply to any research notebook.
Editing & Revisions
Yes. Pencil UI for natural-language edits in NotebookLM. PPTX export for edits in PowerPoint/Google Slides.
Yes. Queue revisions across slides, then Generate Revised Deck for one-pass application.
1 generation + 3-5 targeted revisions. Approximately 5 minutes of revision work.
No. Source-grounding is maintained. The model flags gaps rather than hallucinating.
Old: element-level, one change at a time. New: prompt-based, queue across slides, batch apply. 30 seconds vs 10+ minutes.
No version history yet. Export PPTX before revisions to preserve pre-revision state.
Yes. Convert PPTX to PDF, upload as source. NotebookLM generates a restructured version.
Export
PPTX (editable) and PDF. Google Slides native export coming soon.
Not yet. Export as PPTX, then open in Google Slides. Takes about 30 seconds.
Strategy
A deck without a brief is a tour of sources. A deck from a brief is a weapon for a specific room. The framework adds 10 min upstream, saves 3 hours downstream.
Trinity Engine for executive-grade audiences. Plain Instant Generation for internal review, team updates, course modules.
Decision: audience commits in the room. Persuasion: audience acts later. Persuasion decks need commitment devices on the closing slide.
Build separate decks. Same notebook, three briefs, three generations.
It overlaps. Maps to McKinsey's pyramid principle and SCQA framework, adapted for NotebookLM's citation discipline.
Advanced Use Cases
Yes. Add each URL individually. NotebookLM auto-transcribes. Generate deck from combined transcripts.
Answer-first: recommendation on slide 1, supporting arguments next, evidence last. Each slide opens with a conclusion headline.
Generate a VSL-structured deck. Screen-record with voiceover using Loom, OBS, or Zoom. Ready in under 3 hours.
Upload business plan + market research + financials. Use a pitch prompt: 10-14 slides, problem-solution-traction-market-team-ask.
Yes. Upload PDFs as sources. Synthesize findings across multiple papers. Works with papers up to 300 pages each.
Comparisons
NotebookLM: source-grounded with citations. Gamma/Beautiful.ai: design-first without source grounding. Use NotebookLM when accuracy matters.
Yes. Supports multiple languages. Highest quality for EN, ES, FR, DE, JA, PT.
Sharing & Limits
Yes. Share the notebook or export as PPTX/PDF.
Free: 8-15 slides. Plus: up to 20. AI Ultra: up to 25. No limit on decks per notebook.
Yes. The Boardroom Brief Generator runs against your source notebook, pulling evidence from your actual sources.
Wait for source analysis to complete. Convert unsupported formats (PPTX, Excel) to PDF. Reduce source count if over quota.
30 seconds to 10+ minutes depending on source complexity. Basic (1-3 sources): 30-60s. Standard (3-8 sources): 60-90s. Complex (10+ sources): 2-3 min. Heavy (100+ sources): 15-30 min.
Yes. The 2026 Pencil UI accepts natural-language revision prompts. Queue across slides, batch-apply in one pass. PPTX export produces fully editable text boxes in PowerPoint and Google Slides.
A 7-section strategic brief that runs against your source notebook before any slides are generated. Defines: decision, audience tension, evidence arsenal, 7-slide spine, opening move, closing close, pre-mortem. Adds 10 minutes upstream, saves 3 hours downstream.
3 agents debate each slide: the Advocate makes the strongest case, the Skeptic attacks it, the Synthesizer produces a balanced final with confidence scores. Catches errors single-agent misses. Average deck confidence: 0.74.
No. Content is generated exclusively from your uploaded sources using RAG. Every claim traces to a specific source passage.
Yes. The multi-agent pipeline handles 50+ papers: Source Curator ingests and credibility-scores, Research Architect detects gaps, 3-agent debate produces synthesized slides with confidence scores.
In testing across 80+ cycles: 1 generation + 3-5 targeted revisions. Plan for ~5 minutes of revision work. Confidence scoring tells you when to stop.
PPTX (editable, new in 2026) and PDF (fixed layout). PPTX opens cleanly in PowerPoint and Google Slides. Google Slides native export coming soon.
It overlaps. The "decision being asked for" maps to McKinsey's pyramid principle. The "audience tension" maps to SCQA. Adapted for NotebookLM's grounded-generation strengths — citation discipline that freestyle builders lack.
Yes. Add each YouTube URL individually (auto-transcribed). Generate a curriculum deck from combined transcripts. Works well for lecture series → structured slide sets.
An adversarial stress-testing system. 3 agents generate competing versions of each deck, attack each other's versions, and produce a synthesis better than any single version. Use after the Round Table builds the initial deck.
NotebookLM when you need source-grounded slides with citation trails. Claude for nuanced debate and long-context synthesis. ChatGPT for structured deliverables via Custom GPTs. Grok for speed and unfiltered critique.
30 Killer Prompts for Slide Deck Mastery
6 free prompts above. 30 more below — organized by workflow phase. Each tested across 200+ iterations.