Verdict: Pairing Google's free Gemini Notebook (the renamed NotebookLM) with a local Obsidian vault solves the single biggest problem with AI tools — they forget you the moment you close the tab. By feeding your notes, projects, and customer context into a shared vault that every AI agent reads from, you stop re-explaining your business every morning and start building on yesterday's work. The setup costs nothing, takes about 30 minutes, and compounds daily.
Last verified: 2026-07-30 TL;DR: Gemini Notebook is a free Google research tool that turns any source (PDF, website, video, doc) into 12+ output formats — audio shows, videos, slides, flashcards, mind maps, infographics, and more. Obsidian is a free local notes app that stores everything as plain Markdown files. Wire them together and your AI finally remembers you.
- Best for: solo operators, small teams, and builders who use AI daily
- Cost: $0 for both core tools (Gemini Notebook free tier + Obsidian free)
- Setup time: ~30 minutes
- Pricing/availability changes often — re-check before relying on specific limits.
What Is Gemini Notebook (Formerly NotebookLM)?
Gemini Notebook is a free, source-grounded AI research tool from Google. On July 16, 2026, Google renamed NotebookLM to Gemini Notebook — same product, same notebooks, same address at notebooklm.google, just a new name under the unified Gemini brand. Existing links redirect automatically and no admin action is required for Workspace customers.
In plain terms: you create a notebook (a container), drop in sources (a PDF, a website, a Google Doc, a YouTube link, an audio file), and the tool reads and understands all of it. Then it turns that material into over a dozen different outputs — audio overviews, video overviews, slide decks, mind maps, flashcards, quizzes, infographics, data tables, and downloadable reports. Google's June 8, 2026 upgrade post confirmed the product now runs on Gemini 3.5 with a new agentic framework called Antigravity, adding a secure cloud computer that can write and run code directly inside your notebook.
More than 30 million people already use the tool, according to Google's own announcement — but most use it as a one-shot summarizer and miss 90% of what it can do.
How much does Gemini Notebook cost?
The free Standard tier costs $0/month and includes 100 notebooks, 50 sources per notebook, 50 daily chat queries, 3 audio overviews, 3 video overviews, and 10 Deep Research runs per month — enough to test the full workflow described in this article. Paid tiers bundle into Google's AI subscriptions: AI Plus at $4.99/month (200 notebooks, 100 sources, 200 daily chats), AI Pro at $19.99/month (500 notebooks, 300 sources, 500 daily chats), and AI Ultra at $99.99–$199.99/month (500+ sources, 2,500–5,000 daily chats). These figures come from Google's Gemini Notebook upgrade help documentation and were confirmed live in July 2026. The free tier is genuinely usable — the caps limit volume, not features.
What Is Obsidian and Why Does It Matter for AI?
Obsidian is a free, local-first note-taking app that stores every note as a plain-text Markdown file in a folder on your computer (Obsidian calls this folder a "vault"). There is no cloud database, no proprietary format, and no vendor lock-in — you can open any note in a basic text editor on any device, forever. Obsidian's pricing page confirms the core app is free for personal and commercial use, with optional paid add-ons for Sync ($4–8/month) and Publish ($8/month).
The reason Obsidian matters specifically for AI is simple: because notes are plain text files, any AI agent can read, write, and search them directly. You do not need an API, a database, or a cloud service. If your agent can read a file on disk, it can read your vault. That is what makes it the ideal memory layer.
This is not a theoretical claim — open-source projects like the ai-memory-vault repository (398+ stars as of July 2026) and the Obsidian AI Agents plugin are built entirely around this idea: turn an Obsidian vault into persistent, cross-session memory for AI tools that otherwise forget everything.
Why Pair Gemini Notebook With Obsidian?
The core problem Gemini Notebook solves — turning raw sources into structured content — is powerful but incomplete on its own. Every time you open a fresh notebook, the AI starts from zero. It does not know your business, your brand voice, your last project, or your customers. You explain yourself over and over.
The fix is a memory loop:
- Obsidian holds your memory. Your vault contains your goals, projects, client notes, ideas, and journal — one folder, plain text.
- Gemini Notebook processes information. You drop in a source (a customer call transcript, 50 research articles, a competitor's PDF), and it produces structured outputs.
- The output feeds back into your vault. The summaries, mind maps, and reports get saved as notes.
- Every AI agent reads the same vault. Whatever one agent learns becomes context for all of them.
The system gets smarter every day because you keep feeding it. It never forgets. It never starts over. This is the difference between using a tool and operating a system — and it is the core idea behind what we have previously called building an AI agent operating system.
How to Set Up the Gemini Notebook + Obsidian Memory Loop
Step 1: Create your Obsidian vault
Download Obsidian (free, macOS/Windows/Linux/iOS/Android). Create a new vault — a folder on your computer. Inside it, add a few notes:
Goals.md— your top 3–5 goals for the quarterProjects.md— what you are actively buildingClients.md— customer notes, pain points, winsBrand Voice.md— how you talk, words you use, words you avoidJournal— a daily note folder for raw thinking
Keep it simple. The vault does not need to be perfect — it needs to be honest and current.
Step 2: Create a Gemini Notebook and add your vault as context
Go to notebooklm.google and sign in with any Google account. Click New notebook. In the Sources panel, click Add and upload your key Obsidian notes as sources — export them as Markdown or PDF, or paste the text. Now the notebook knows your goals, your context, and your voice before it generates anything.
If you want to go deeper, you can use Discover sources (Deep Research) to have Gemini Notebook find and add web sources for a research question, then layer your vault notes on top for grounding. Free accounts get 10 Deep Research runs per month.
Step 3: Feed in a real source and generate outputs
Drop in a source you want to work with: a long customer interview transcript, a batch of research PDFs, a YouTube link, a Google Doc. Then open the Studio panel and generate outputs. As of the June 2026 upgrade, Google's official announcement confirmed the output formats now include:
| Output Format | What It Produces | Best For |
|---|---|---|
| Audio Overview | A conversational podcast-style audio show | Commute review, repurposing |
| Video Overview | A narrated slideshow video | Social media, course previews |
| Mind Map | A visual hierarchy of themes and connections | Grasping a new topic fast |
| Flashcards | Q&A cards with progress tracking | Learning, onboarding |
| Quiz | Comprehension test with tracking | Team training |
| Slide Deck | Editable slides (exportable to PPTX) | Presentations |
| Infographic | A visual summary graphic | Lead magnets, social posts |
| Data Table | Structured comparison table (export to Sheets) | Literature reviews, product compare |
| Report | Downloadable PDF with charts and tables | Client deliverables |
All outputs are grounded in the sources you uploaded — and because your vault notes are in there too, the outputs carry your context and brand voice.
Step 4: Save the output back to your vault
Take the summaries, mind maps, and reports Gemini Notebook generates and save them as new notes in your Obsidian vault. The work you did today becomes context for tomorrow. After a month of this loop, your agent has a mountain of accumulated knowledge it draws on automatically — covering areas from building a personal AI agent operating system to organizing your content production pipeline.
Step 5: Wire your vault into your AI agent
This is where the compounding effect kicks in. If you use an AI agent platform that can read local files (Claude Code, Codex, Hermes Agent, or any tool that accesses your filesystem), point it at your Obsidian vault directory. Now every agent reads the same notes — what one learns, they all learn. They share one brain.
We have written a full guide on how to use Claude with Obsidian as a free AI second brain, and another on building a free agent OS with Hermes and Obsidian. The same vault that feeds Gemini Notebook feeds those agents — no duplication.
6 Practical Business Use Cases for the Loop
1. Content multiplication from one recording
Drop a podcast or customer interview into Gemini Notebook. In a few clicks, generate an audio clip, a narrated video, a slide deck, and an infographic — all from that single source. Because your vault already knows your brand voice, the outputs sound like you, not a generic AI. One recording becomes a week of content.
2. An SEO content engine
Feed your case studies, research notes, and product docs into a notebook. Ask it to generate article drafts grounded in those sources. Those articles help people find you on Google and in AI search results. Because the outputs are grounded in your real work, they carry genuine information gain — not generic filler.
3. Lead magnets at scale
Take one guide or whitepaper and turn it into a video, an audio show, and an infographic. Now you have a free value asset deployed across multiple formats and channels — all from a single source document.
4. Sales call summaries for your team
Upload a customer call recording. Gemini Notebook produces a clean one-page summary your team can actually use. New team members get up to speed on accounts fast, and the summary feeds back into your vault so future conversations have context.
5. Course creation while you sleep
Upload what you know — notes, transcripts, existing content. Gemini Notebook builds slides and flashcards for every lesson. You are assembling a course module-by-module without writing each one from scratch. Save the slide decks back to your vault as templates.
6. Research synthesis in minutes instead of hours
Feed it 50 articles on a topic. It gives you one mind map showing how everything connects, plus a structured data table comparing key claims across sources. What used to be 40 hours of reading becomes a scan of a visual map and a downloadable table.
If you want to go further with this pattern, our Gemini Notebook complete update guide covers the full feature set and recent changes, and our guide on organizing Gemini Notebook collections shows how to structure multi-notebook research.
Gemini Notebook vs Obsidian: What Each Tool Does
| Dimension | Gemini Notebook | Obsidian |
|---|---|---|
| What it does | Reads sources, generates structured outputs | Stores and links your notes locally |
| Where it lives | Browser (web app at notebooklm.google) | Desktop app on your computer |
| Data location | Google cloud | Your local filesystem |
| Persistence | Each notebook starts fresh unless you add context | Everything stays — it is your permanent memory |
| Cost | Free tier ($0) to Ultra ($99.99+/mo) | Free core app; Sync $4–8/mo, Publish $8/mo |
| AI capability | Built-in Gemini 3.5, source-grounded | No built-in AI — but any agent can read/write the vault |
They are complementary, not competitive. Gemini Notebook is the content engine; Obsidian is the memory. The loop makes them a system.
What This Means for You
If you use AI tools daily and still start every session by pasting the same context into a fresh chat, you are burning time. The people who set up a real memory system this year will pull ahead because their AI gets smarter every day — it knows their goals, their customers, their voice, and their last decision. The gap between a system user and a tool user compounds, and it compounds fast.
Start small: make one Gemini Notebook today, drop in five things about your business, and watch it turn them into a video and an audio show. See the output with your own eyes. When you are ready to plug it into your memory vault and make every agent know you, that is the whole system — and it costs $0 to start.
If you are already running an AI agent operating system, the Gemini Notebook loop slots in as your content-generation layer — your vault remains the single source of truth.
FAQ
Q: Is Gemini Notebook free to use?
A: Yes. The free Standard tier costs $0/month and includes 100 notebooks, 50 sources per notebook, 50 daily chat queries, 3 audio overviews, 3 video overviews, and 10 Deep Research runs per month. No credit card or trial is required. Source: Google's Gemini Notebook upgrade help doc.
Q: Is Gemini Notebook the same as NotebookLM?
A: Yes. Google renamed NotebookLM to Gemini Notebook on July 16, 2026. Same product, same notebooks, same URL (notebooklm.google). Existing links redirect automatically. Source: Google Workspace Updates.
Q: Do I need to know how to code to set up the Obsidian memory loop?
A: No. Obsidian is a desktop app — you install it and type notes. Gemini Notebook runs in your browser with point-and-click buttons. The "loop" is simply saving Gemini Notebook outputs as notes in your vault folder. No code is required for the core setup.
Q: How many sources can I put in one Gemini Notebook on the free tier?
A: 50 sources per notebook on the free Standard tier. The AI Pro tier ($19.99/month) raises that to 300 sources, and AI Ultra ($99.99+/month) goes to 500–600. Source: Google support documentation.
Q: Can Obsidian really work as AI memory without a database?
A: Yes. Obsidian stores notes as plain Markdown files in a local folder. Any AI agent that can read files on disk can read, search, and write to your vault. No vector database, API, or cloud service is needed — though some setups add one for semantic search across very large vaults. Source: Obsidian pricing and features.
Q: What is the difference between Gemini Notebook and the Gemini app?
A: Gemini Notebook grounds every answer only in the sources you upload (or that its Deep Research feature finds and you approve), which keeps responses citable and honest. The Gemini app can use web search and general knowledge broadly. Use Notebook when you want answers tied to specific documents; use the Gemini app for open-ended work. Studio outputs (audio, video, slides) are features of Gemini Notebook specifically.
Discussion
0 comments