Research OS for NotebookLMLocal-first captureAnalytics + source viewsAutomations + exports
Kortex
NotebookLMChatGPTClaudeGeminiPerplexityRedditLinkedInX / Twitter

Turn NotebookLM into a searchable research operating system.

Kortex connects AI chats, notebook captures, imports, prompts, automations, analytics, source views, and exports into one durable knowledge workflow instead of a pile of isolated sessions.

8+Capture surfaces
9Automation recipes
19Dashboard sections
Local-first + hosted syncStorage model
Live workspace shell

Save chats, sources, prompts, tags, and exports into one operator console.

Bulk source controls, notebook favorites, source views, analytics, and cleaner public-site flows make the clone behave more like a real product.

ManagementAnalytics, tags, favorites
Automation9 recipe patterns
ExportsMarkdown, ZIP, Text
Why operators outgrow raw NotebookLM fastCapture is only the first step. Search, recall, structure, and repeatability are the real bottlenecks.

Why it exists

NotebookLM is strong for synthesis. Kortex adds the operating layer around it.

Without Kortex

Research gets trapped in isolated notebooks.

Constraint

Chats, sources, and exports stay trapped inside separate notebooks and browser tabs.

Constraint

Knowledge gets messy once the notebook count grows past a handful of projects.

Constraint

Imports, exports, and source cleanup become a repetitive click-heavy process.

Constraint

Insights stay passive instead of becoming reusable operating procedures.

With Kortex

You get a reusable system instead of a one-off session.

Upgrade

Capture conversations, source snapshots, prompts, and artifacts into one searchable workspace.

Upgrade

Use nested folders, favorites, tags, collections, and saved source views to keep research structured.

Upgrade

Run bulk imports, bulk source moves, notebook merge flows, and cleaner export pipelines.

Upgrade

Save prompts, auto-tag captures, trigger automations, and turn notebooks into podcast feeds.

Feature map

The live clone now covers the broader solo-research surface much more closely.

Capture

Collect from the tools you already use

NotebookLM overlays, chat adapters, social captures, web highlights, RSS, CSV, tabs, and Google Docs imports all feed the same system.

Organize

Treat research like an operating system

Folders, notebooks, favorites, tags, prompt folders, collections, source views, and merge / move flows keep knowledge tidy instead of temporary.

Automate

Turn saved context into repeatable output

Smart sorters, generators, pipelines, cleaners, podcast curation, and auto-research recipes help new captures keep moving.

Search

Find the right note, source, prompt, or artifact fast

Search spans notebooks, prompts, collections, podcasts, tags, and, on hosted plans, retained sources and artifacts.

Export

Get clean outputs instead of messy copy-paste

Preview and export notebooks or artifacts as Markdown, text, ZIP bundles, and print-friendly HTML.

Operate

Stay local-first or unlock hosted controls

The stack runs locally, supports hosted sync, and already includes enterprise controls such as SSO posture, provisioning switches, and zero-retention options.

Inside the product

What changed in the workspace surface

Workspace analytics

Track notebook volume, favorite density, source-type mix, top tags, collection coverage, and recent activity from one dashboard.

Source management

Search, filter, paginate, bulk select, move, delete, and spin a new notebook out of selected sources in one pass.

Tag operations

Create, rename, merge, and delete workspace tags globally instead of editing them notebook by notebook.

Saved source views

Store reusable source filters by notebook, query, tag, source type, and time window so recall is faster.

Prompt library

Keep prompts in folders with categories and tags, then reuse them across research loops instead of retyping from memory.

Bulk output

Generate artifacts in batches, export notebooks cleanly, and publish podcast feeds from curated notebook context.

Operator proof

Why this feels more like a serious product now

Bulk source cleanup, favorites, tags, and exports finally feel like one coherent workflow instead of six separate chores.

Research workflow operator

Saved source views and the prompt library make this feel much closer to a real knowledge operating system than a notebook dump.

NotebookLM power user

The live dashboard now covers the surfaces I would expect from a serious clone: analytics, tagging, bulk management, and cleaner navigation.

Solo analyst

Built for serious research workflows

OpenAIGoogleMicrosoftMetaNVIDIAHarvardYalePrincetonColumbiaCornellAnthropicAmazon

Pricing

Choose the layer that matches how serious your knowledge workflow is.

Free

$0/ month

Local-first workflows for individual research.

  • Local storage only
  • Prompt folders and saved views with limits
  • Core capture surfaces and exports
  • No hosted sync or automation pipelines
Start free

Lifetime Pro

$69 one-time

One-time unlock for long-horizon research systems.

  • Everything in Pro
  • One payment instead of a recurring subscription
  • Best fit for heavy solo researchers
Get lifetime access

Teams

$12/ seat / month

Small-team collaboration with stronger control and shared operating patterns.

  • Shared prompt libraries
  • Admin-facing controls and usage visibility
  • Team notebook operating patterns
  • Better fit for labs and small research pods
Explore teams

Enterprise

Custom

Security posture, SSO controls, and centralized operating controls.

  • Centralized billing and admin controls
  • SSO / provisioning posture
  • Zero-retention and server-retention options
  • Organization-level governance settings
Contact sales
Instant value

One click can save a chat, a source, a highlight, or an artifact into a structured notebook.

Better recall

Search, analytics, source views, and reusable prompts make past work useful instead of buried.

Less admin drag

Bulk imports, source move flows, favorites, tag management, exports, and automations reduce manual cleanup.

FAQ

What is complete now, and what still depends on provider credentials?

What is still pending for full live parity?

Core app flows are live. The remaining provider-backed pieces are Firebase auth, Google Docs OAuth, and Dodo billing.

What can I capture into Kortex?

The extension and dashboard cover NotebookLM, ChatGPT, Claude, Gemini, Perplexity, Reddit, LinkedIn, X/Twitter, and general web highlighting through optional permissions.

Is this local-only or cloud-enabled?

You can keep it local-first, or on hosted plans use the dashboard and sync APIs for retained entities, exports, automations, and cross-device continuity.

How close is this clone to the public Kortex product?

The clone now covers the broader solo-research surface much more closely, including analytics, favorites, tags, source management, source views, onboarding, shortcuts, and a fuller public site shell.

Ready to operate your knowledge instead of renting it?

Use the live dashboard now, then wire the provider-backed layers when you are ready.