Documents and notes decision
Kortex-Notebooklm
A focused subset (capture from Gemini, searchable storage, basic RAG summaries, and TTS exports) is realistic for one capable developer in ~30 hours; fully reproducing the team's breadth of integrations, automation UI, polished cross-device sync, and team/SSO features is much larger and likely needs more engineering and ops.
Visit website↗Open-source builds that already do this
Every project below is open source and already does this job today. Fork one, self-host it, or take the parts you need — the build prompt further down assumes an empty file, and this is the shortcut past that. Licences differ; check the one on each card before you ship.
What a replacement has to do
- Capture chats/sources from Gemini Notebook (or web), persist them in a searchable store, show/edit/highlight sources in a dashboard, run simple retrieval for question/summary and export artifacts (PDF/MD/audio).
What it still won’t have
- Polished multi-language UI, weekly feature cadence, and UX polish
- Pre-built integrations (Google Docs sync, many third-party imports) and team/SSO admin tooling
- Automation pipeline UI with pre-built rules and hosted reliability/uptime SLAs
- Podcast feed management and hosted audio bandwidth at scale
What remains hard
- Product polish and ongoing maintenance
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 15 seats.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a minimal self-hosted replacement for Kortex consisting of a Chrome extension + single-node web app. Stack: React (TypeScript) frontend, Express.js backend (Node 18+), Postgres (or SQLite for single-user) with FTS, and a simple task worker (BullMQ) for background audio generation. In scope: (1) Chrome extension that extracts Gemini Notebook chat HTML and POSTs normalized source objects to the backend; (2) backend endpoints to create/list/update/delete sources, notebooks, collections, and tags; (3) full-text search and a retrieval endpoint that returns the top-k sources for a query and a summarization endpoint that calls a configurable LLM API (OpenAI/Anthropic) to produce an artifact; (4) web dashboard to browse/edit sources, create notebooks, run exports (Markdown/ZIP), and trigger audio generation; (5) TTS integration using a configurable provider (e.g., ElevenLabs or open-source TTS) and serve downloadable MP3s; (6) basic auth (email+password) and a single-user session. Out of scope: multi-tenant billing, SSO/enterprise provisioning, extensive import adapters (only implement web page capture and CSV import), advanced pipeline UI with many prebuilt rules, and mobile apps. Require: error handling for failed imports/LLM calls, unit tests for API endpoints and core parsing logic, and Dockerfiles + a docker-compose dev/start script for local deployment.
How we checked
How the score was reached
- Partly verdict base52
- An open-source build was found+5
- 4 cited sources+3
- Price verified on pricing page+3
- Evidence score63
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time — so the same evidence always produces the same number.
How scoring works →Cited sources · 4
Every page the run actually retrieved.
- official productKortex - Transform Gemini Notebook into Your Knowledge Hub (homepage, includes pricing)
- official docsKortex Features
- open sourcelabring/FastGPT
- open sourcenashsu/llm_wiki
Integrity checks
What held up, and what did not.
