reflect vs innernet
in reflect, you draw the graph. in innernet, the graph gets drawn from your work.
the short version
In reflect, you draw the graph. In innernet, the graph gets drawn from your work.
reflect is a fast, beautiful, end-to-end encrypted notes app with backlinks and a native AI assistant — the best-executed version of the networked-thought idea. Its promise is that it "mirrors the way you think." It does, on one condition: that you keep writing the notes and keep making the links.
innernet has no notes and no links you maintain. netti reads what accumulates and writes the structure itself, then keeps it current as your work changes.
what reflect is building
A second brain. Networked notes with backlinks that form an idea graph, end-to-end encryption, sync across desktop, iOS and web with offline support, calendar integration for Google and Outlook, a web clipper, Kindle highlight sync, and an AI assistant on GPT-4 and Whisper that transcribes voice, drafts outlines, summarises meeting notes, fixes grammar and runs your saved prompts. One-click publishing. $10 a month billed annually.
It is fast and it is well made. For people who genuinely write to think, it's one of the best tools available.
what innernet is building
A personal memory layer under the AI tools you already use. npx innernet connects Claude Code, Cursor, Codex, Windsurf, Gemini CLI, Claude Desktop, VS Code, Cline, Zed and Continue to the same memory, with an MCP endpoint for anything else. netti — one per person — files what accumulates into Context Maps: dimensions that read like living documents, nodes for the people and decisions that recur, commits and branches you can work inside.
The unit is not a note. It's a conclusion with its history attached.
where they diverge
1. maintained structure vs. emergent structure
A backlink graph is a map you draw and have to keep drawing. It's an honest system — the structure is exactly as good as your discipline, which is why second brains decay when people get busy. innernet's dimensions are written by netti from what accumulates and updated when the work contradicts them. The structure survives the week you had no time for it.
2. AI in the tool vs. memory for every tool
reflect's assistant makes you better at working inside reflect — transcribe, outline, summarise, tidy. innernet doesn't have an assistant you use; it makes every assistant you already use behave like it knows your project. The difference shows up in Cursor at 11pm, where reflect's AI isn't and your context is.
3. what you wrote down vs. what you did
Notes contain what you thought to record, and the honest truth is that most decisions never get written down — they happen in a conversation with a model, in a commit message, in a rethink at 1am. innernet reads that residue. A notes app can only ever be as complete as your habit of using it.
side by side
| reflect | innernet | |
|---|---|---|
| Unit | a note | a fact or decision, with sources and history |
| Who builds the structure | you, via backlinks | netti, from what accumulates |
| AI's role | assistant inside the app | makes your other AI tools context-aware |
| Reachable by Claude Code / Cursor | no | yes — one command |
| Encryption | end-to-end | optional column-level, per project, behind your passphrase |
| Change over time | note edits and history | versioned positions, branches you can park and merge |
| Pricing | $10/mo billed annually | free to start, no card |
where reflect is genuinely better
- End-to-end encryption. Stronger than our default posture and a legitimate reason to choose it.
- Writing experience. Fast, minimal, pleasant. If you think by writing, that matters more than any feature list.
- Capture surfaces we don't have. Kindle highlights, web clipper, calendar integration.
- You can see everything. A notes app is fully legible — your notes are right there, in your words. Nothing is inferred.
- It's finished. A mature, stable product with a clear price.
where innernet is stronger
- It doesn't depend on your discipline. The memory forms whether or not you had a good week.
- It's read where you work. Your context arrives in the editor and the terminal, not in a separate tab.
- Positions are versioned, not just notes. Branch a direction, park it, roll it back — what changed is as retrievable as what's current.
- Disclosure per fact. Private, self, trusted, work, public, enforced when a tool reads.
- netti is doing editorial work. Deciding what earns a dimension, updating it when the work moves, noticing the node that keeps recurring.
which one you want
Use reflect if you write to think and you want the best place to do it, with your own hands on the structure.
Use innernet if the memory you need isn't notes you'd write — it's the reasoning already scattered through your work, and you want six AI tools to read the same version of it.
Both is very reasonable. reflect is where you think on purpose; innernet holds what you thought while doing something else.
the question we actually get
"My notes are mine and I can read every word. Yours are written by an agent. How do I trust that?"
By checking it, which is why we don't hide it. Context Maps are readable documents, not embeddings — you can open a dimension and see what netti concluded, in prose, with a commit history showing when it changed and what it replaced. If it's wrong, you say so and it changes, and the previous version stays in the history. A memory you can't audit isn't one we'd use either.
Accurate as of 3 September 2026, based on reflect's public site. If we've described their product incorrectly, tell us and we'll correct it.
this is what innernet remembers. you can keep your own.
innernet is memory your ai tools read live — projects, decisions, the things you’d otherwise re-explain every session. we write these letters out of ours.