1. NotebookLM — the research notebook
What it is: Google's research assistant that ingests up to a few dozen sources per notebook and answers grounded questions about them, with citations back to the source.
Price: Free (Plus features via Google One AI tiers).
Best for: One-off research projects with a fixed source set — a paper, a legal brief, a due-diligence dossier.
Key limitation: Google-only, notebook-scoped. No MCP surface, no path into a cross-tool workflow. It's a superb single-purpose tool, not a lifelong knowledge base.
2. Mem — the AI notes app
What it is: A note-taking app built around organising and surfacing what you write with AI.
Price: Around $10/mo.
Best for: People who write a lot and want automatic organisation without folders or manual tagging.
Key limitation: Capture is manual — if you don't type it in, it's not in there. The AI surface is Mem's own; portability across other AI clients isn't the product bet.
3. Reflect — networked daily notes
What it is: A fast, encrypted, backlinked daily-notes tool with an integrated GPT-style assistant.
Price: $10/mo annual.
Best for: Journaling and light knowledge work with an assistant on tap inside the notes app.
Key limitation: Same shape as Mem on capture — manual input — and the assistant is scoped to Reflect. If you consume more than you write, this isn't the wedge.
4. Readwise — the highlight system of record
What it is: A spaced-repetition highlight tool that aggregates from Kindle, Instapaper, articles and (via Reader) the modern web.
Price: Roughly $8–10/mo.
Best for: Daily highlight review that keeps books and articles alive in your head over months and years.
Key limitation: Highlights, not corpora. Native video and audio understanding is minimal, retrieval is shallow, and MCP isn't the shape of the product. Great alongside a real knowledge base; insufficient as one.
5. Fabric — the visual workspace
What it is: A visual, canvas-style workspace for saving files, links and multimedia into an organised, AI-searchable library.
Price: See site — public pricing has moved during 2026; verify before committing.
Best for: People whose knowledge is visual — moodboards, references, image-heavy research — and who want a spatial UI rather than a list of pages.
Key limitation: Visual-first also means text-heavy long-form (transcripts, PDFs of research papers) is not where the product is strongest. AI is in-app; MCP is not the focus.
6. Obsidian — local-first and yours forever
What it is: A Markdown editor over your own file system with a plugin ecosystem the size of a small planet.
Price: Free (Sync $4/mo, Publish $8/mo).
Best for: Long-lived personal notes you enjoy tending, and anyone whose top priority is 'the files are mine, on my disk, forever'.
Key limitation: The maintainer is you. Backlinks, tags, embeds and plugin conflicts are only ever as coherent as the last evening you spent gardening the vault. AI features exist through third-party plugins of uneven quality; a first-party AI-agnostic bridge is still community work.
7. BrainTube — AI-agnostic memory layer (this is us)
What it is: A persistent memory layer for AI. It auto-captures YouTube videos, podcasts, articles and PDFs, transcribes them, builds a knowledge graph, and exposes the whole library over a native MCP server to Claude, ChatGPT (via connectors), Cursor, Gemini and any other MCP-compatible client. Recall cited to the exact timestamp in the source. Chrome extension for capture. Full JSON export on every plan, including Free.
Price: Free $0 (30 credits/mo) · Starter $9 (250) · Pro $19 (1,000) · Pro+ $49 (5,000).
Best for: People who consume a lot of long-form video and audio and want the same library queryable from every AI tool they use — including AI tools that don't exist yet.
Key limitation, honestly stated: A younger product than Notion, Obsidian, Recall or Readwise, with a smaller community and a smaller plugin surface. In-app reading and summarization UX is functional but less polished than Recall's. If your workflow depends on Notion or Obsidian's ecosystem, keep it — BrainTube is a memory layer for AI, not a workspace replacement. Direct head-to-head with Recall at /compare/braintube-vs-recall.
How to actually pick
Ask three questions in order. First: does the tool capture what you actually consume without your intervention? If you have to hand-type every source, the tool will lose to your calendar. Second: does the tool expose your knowledge over an open protocol like MCP, so it's usable from any AI client you use now or later? If not, you're renting your memory from one vendor. Third: can you leave with your data intact? If export is gated, absent or partial, treat the corpus as impermanent regardless of the marketing.
The related reading below goes deeper on each of those axes — including what an AI memory layer actually is, how to give your AI persistent memory across every tool you use, and how to chat with your YouTube videos and podcasts using AI.
