Curtext Curtext
Curtext

Context. Always.

A revolution in team AI

Curtext is a business-wide AI work environment that always includes relevant business context, allows you to switch between the latest model providers and lets you collaborate cursor-style on markdown documents. Imagine a multi-user, multi-model Claude Cowork with awesome Obsidian markdown skills.

Positioning refresh

Auto-Commit snapshots every change, so nothing is lost. True real-time CRDT is the deep, later moat, mainly for multiple humans.
So CRDTs are the real-time piece - when do we actually need them?
Only once two humans edit the same file live. Until then, Auto-Commit's turn-based merge covers every case.
Makes sense - park it for Phase 2.
Aside

What does CRDT mean?

Conflict-free Replicated Data Type - lets many people edit the same doc at once without their changes clashing.

Ask a follow-up…

Sharpen our positioning line; make it lead with what we do.
Done. I pulled audience from strategy.md and revised positioning.md - review the doc →

Reply…

… / Brand positioning /  positioning.md

1.2 KB · 12 lines · Syntax visible, human/AI editable

# Positioning

## One-liner

-We help teams use AI better.
+Curtext is the AI workspace that already
+knows your business - context-loaded, not cold.

## Audience

+- Teams of 20–500 already running on AI
v14 · reversible

Features

  • - Chat with all the latest models from Anthropic, OpenAI, Gemini
  • - Twin mode pairs 2 (or more) providers for outsized gains / cross-checks
  • - Cascading Context company, team and project context in every chat
  • - Markdown Scratchpads editable by you, your AI and your team
  • - Asides side-chats for quick questions without poisoning the whole context
  • - Distill compress long chats, but keep a full version history
  • - Auto-merge intelligent markdown merging with Cursor-like accept / decline
  • - Auto-commit never lose anything with git-like automatic versioning

Like Cursor; but for words.

Request early access All Features

Problems → Solutions

Everything wrong with team AI chat today, fixed.

Chats don't have company context by default.

Markdown context Every chat is pre-loaded with your company, team and project markdown as standing context.

Individual team members don't share consistent context.

Markdown drive A shared markdown database means everyone can share and edit team documents.

Claude et al. can't collaborate on documents - only generate, download, then re-upload.

Markdown Scratchpads provide Collaborative Markdown: edit documents together right in the chat window - no download/re-upload loop.

Chats become unwieldy once they pass a certain length.

Tl;dr mode the important work lives alongside, as a versioned markdown summary - so you don't get buried in a swamp of endless chat

Chats get littered with side-questions.

Asides branch side-discussions into post-its that sit around the main thread without cluttering or confusing the main context. Things like "what does this acronym mean?" or "explain why quantum breaks crypto"

Complex topics require more than one chat.

Knowledge Trees link chats with Asides to create a whole tree of knowledge

Non-coding teams can't switch models when one falls short, or another would be better or cheaper.

Switch between every frontier model - or let auto mode pick - right in the chat. No code, no API juggling.

Standard chat can't run "twin" models from different providers to check each other.

Twin mode pairs models from different providers to check each other's work and measurably raise quality.

Standard chat has no versioned document history.

Auto-Commit gives every document a git-like, fully restorable version history.

Multiple AI seats per employee is expensive.

One workspace with shared API access to every model - no stacking per-seat subscriptions.

1 brain
Shared across the whole company
3 platforms
Use in the browser, or download for Windows and macOS
All models
Access all of the latest AI LLMs without individual subscriptions. Combine models for extra power!

Collaborative Markdown, for the whole team

Cursor changed how developers write code: the AI proposes, you dispose, line by line, inside the document itself. Curtext brings that exact loop to prose - your company's knowledge, plans, and decisions, written together in plain markdown.

Cursor is for code

Agents and developers edit the same files, with inline diffs and instant review.

Curtext is for prose

Agents and teams edit the same markdown, with the same propose/dispose loop for everyone

  • Markdown is the lingua franca of AI - readable by humans, native to models, diffable by design.
  • Every document is a shared space: humans type, agents edit, both see the same versioned truth.
  • No silent overwrites. Collaboration, not automation - the human always disposes.

Teams running on AI have no shared brain.

Teams running on AI have no shared brain. Every employee's agent starts cold - ignorant of company practice, role context, and what was decided last week - so people re-explain the business to their tools all day, and answers drift because everyone's context is different.

Generic chat assistants

Don't know your company. Every session starts from zero.

Google Docs / Notion

Know your documents, but aren't agent-native and can't act.

Enterprise RAG search

Read-only Q&A over your docs - not a place to write and edit.

An agent workspace pre-loaded with your org's knowledge.

  • Org structure is the context index - company, role, and project markdown loaded every interaction. Curation, not retrieval.
  • Agents produce editable markdown versions; permissions decide who can change what.
  • Everything is versioned - every agent write is attributed and reversible.
  • The store stays dumb; the brain lives in Curtext.

Asides

Ask a side question without derailing the thread.

Hit an unfamiliar term or a quick tangent mid-conversation? Open an Aside - a little post-it with its own chat box, pinned to the exact message. It inherits everything up to that point, but nothing it says ever leaks back into the main thread.

Context-safe by design
An Aside is a frozen snapshot at its branch point. Tangents never pollute the trunk's context - or your token bill.
Not just a post-it
Open, ask, resolve, close. Save it to leave a small margin marker you can re-open later, or delete it and move on.
Promote when it grows up
A tangent that needs its own tangents has outgrown a post-it. Promote it to a full chat, linked back to where it forked.

Main thread

Auto-Commit snapshots every change, so nothing is ever lost. True real-time CRDT is the deep, later moat, mainly for multiple humans.

Aside · branched here

What does CRDT mean?

Conflict-free Replicated Data Type - the data structure that lets many people edit the same document at once without their changes clashing.

Nothing in the Aside feeds back into the main thread.

Twin Mode: Why combine models

A panel of models beats any model alone.

Curtext lets you fuse models - run several together and reconcile their answers, instead of betting everything on one. The result is consistent: every fusion panel outperforms its strongest member working solo. Twin mode is the simplest case (two role-different models checking each other); larger panels push quality higher still.

Notably, the best fusion of mid-tier models can match or beat a top model running alone - so you don't need the single most expensive model to get top-tier answers.

Fusion panel Solo model Scored on 93 of 100 tasks
Fable 5 + GPT-5.5 (fusion)
69.0%
Opus 4.8 + GPT-5.5 + Gemini 3.1 Pro (fusion)
68.2%
Opus 4.8 + GPT-5.5 (fusion)
67.5%
Opus 4.8 + Opus 4.8 (self-fusion)
65.6%
Claude Fable 5 (solo)
65.4%
Gemini 3 Flash + Kimi K2.6 + DeepSeek V4 Pro (fusion)
64.6%
DeepSeek V4 Pro (solo)
60.4%
GPT-5.5 (solo)
60.0%
Claude Opus 4.8 (solo)
58.6%
Kimi K2.6 (solo)
53.8%
Gemini 3.1 Pro (solo)
45.4%
Gemini 3 Flash (solo)
43.0%
40%45%50%55%60%65%70%

Benchmark: 100-task internal evaluation. Axis truncated to 40–72% to show separation. Orange entries scored on 93 of 100 tasks.

Tl;dr Mode

Never lose the thread on a long chat.

Deep conversations drift. You agree on tone, reject a headline, settle on a structure - then forty messages later you're scrolling back trying to remember what you actually decided. The essentials get buried under back-and-forth.

Tl;dr mode keeps a live background summary: a short markdown digest of the important points in the conversation. It updates as you go - decisions, constraints, open questions - so you always have a current snapshot without re-reading the whole thread.

Always current
The summary rewrites itself after meaningful turns. What mattered five messages ago is folded in or dropped as the conversation moves on.
Background, not another chat
It sits beside the thread as reference markdown - not a second conversation you have to maintain.
Decisions at a glance
Tone, structure, names, constraints: the things you'd put in a brief if you had time to write one after every session.

Main thread

Let's draft the launch email for the beta.

Keep it under 200 words. No hype - we're talking to people who already signed up.

Subject line: "You're in" or something plainer?

Plain. "Early access to Curtext" - save "You're in" for the body.

CTA: download link or waitlist confirmation?

Download link. They already have access.

… 34 more messages …

Tl;dr live

Launch email · beta

  • Under 200 words, no hype
  • Subject: Early access to Curtext
  • CTA: direct download link
  • Tone: plain, respectful of their time

Open: subject line "You're in"?

The tl;dr panel updates as decisions land - no manual note-taking.

Download

Every platform. One workspace.

Use Curtext in the browser, or install the native app on your phone or desktop. Same context, same markdown, same propose/dispose loop - wherever you work.

Desktop and mobile apps are in early access - buttons are placeholders for now.

Pricing

Enterprise

Let's talk custom

For larger orgs needing SSO, granular permissions, and twin mode at scale.

  • SSO + MLAuth paired identity
  • Granular per-file permissions
  • Twin mode + real-time co-editing
  • Priority support & SLAs

Indicative pricing - not yet live.

Give your whole company one brain.

We are onboarding a small group of beta partners now. If your team already runs on AI and is tired of re-explaining the business every morning, let's talk.

Request access Read the vision