Manuscript Get started

For researchers, students, and serious writers

Made for Research and Learning

Grounded in research papers from Semantic Scholar, OpenAlex, and PubMed.

The collaborative editor that streams Claude, Gemini, and GPT into the same document — with verified citations, LaTeX math, and exports to Word and PDF.

No credit card · Cancel anytime

Grounded in real sources

Citations are verified against the same databases university libraries use. We don't make references up — we look them up, fetch the PDF when it's open access, and cite the canonical metadata.

Semantic Scholar

Primary discovery layer across 200M+ papers. Powers Deep Research's first-pass paper sweep.

OpenAlex

250M+ scholarly works, open metadata. Cross-checked against Semantic Scholar for coverage in niche fields.

PubMed

NIH's biomedical and life-sciences index. Automatically engaged when Deep Research classifies a query as medical or biological.

PubMed Central

Open-access full-text archive linked to PubMed. Where we pull the actual PDF when the paper is OA.

Crossref

The authoritative DOI resolver. Every citation is verified against Crossref's canonical metadata before it lands in your bibliography.

Unpaywall

Open-access PDF discovery across 50M+ articles. Finds a legal open version when the primary URL is paywalled.

arXiv

Preprint server for physics, math, CS, and statistics. Resolved by DOI as an Unpaywall fallback so the PDF reaches your bibliography.

Wikimedia Commons

Diagrams and figures with explicit Creative Commons licenses. Sourced by Study Guides with author + license cited inline.

Every fetched PDF is attached to your document — no broken links three months later. License and attribution are rendered next to every borrowed figure. The bibliography is renumbered automatically when you cite a new paper.

Everything you need to research, write, and ship

One editing surface for prose, math, code, tables, citations, images, audio, and video.

Multi-model AI

  • Claude Sonnet / Opus, Gemini 2.5 Pro / Flash, GPT-4o / 5
  • Multi-model Send: fan out + synthesize in one pass
  • Smart Send: auto-picks doc context via a cheap router
  • Context clips for text, files, images, YouTube links

Math & science

  • Inline & block LaTeX with live KaTeX preview
  • Auto-numbered block equations
  • Smart paste from Word, ChatGPT, Claude.ai (OMML / MathML)
  • Native math export to DOCX — editable in Word

Citations & references

  • BibTeX, RIS, DOI lookup, manual entry, or free-text
  • APA / MLA / Chicago, auto-renumbering in-text marks
  • PDF attachment per citation
  • Cross-document citation search & reuse

Deep Research

  • Multi-phase pipeline: classify → discover → verify → synthesize
  • Sources: Semantic Scholar, OpenAlex, PubMed
  • Live progress, cost-budgeted ($1–2 / session)
  • DOCX + PDF report exported & emailed when done

Real-time collaboration

  • Y.js CRDT — conflict-free multi-cursor editing
  • Share by email, copy-to-my-docs forking
  • Read-only viewers, owner manages grants
  • WebSocket notifications when peers act

Resources panel

  • Inline viewers for PDF, DOCX, XLSX, Markdown, code
  • Drag from panel into the doc as cards or AI clips
  • Common files shared across all your docs
  • Browser extension captures web pages into Resources

Audio & voice

  • Record audio in-browser, save as attachment
  • One-click Whisper-1 transcription
  • Transcripts saved as linked .txt files
  • Gemini audio-capable models accept the binary directly

Built to ship

  • Auto-save, snapshots with word-level diff
  • DOCX with editable math, PDF with KaTeX, native .owdoc
  • Cancel mid-stream, recover after disconnect
  • Mobile-responsive from 320 px, dark mode default

How we compare

The capabilities that aren't available together anywhere else.

Feature Manuscript Notion AI Overleaf Lex SciSpace
Choose your LLM per request multi-model
Fan-out to 3 providers + synthesize
AI streams into the doc side panel side panel inline edits
Native math in DOCX export partial
Citation manager built-in
Deep Research pipeline
Image + video + music generation partial
Audio recording + transcription
Real-time multi-cursor collab partial

✓ = fully supported · partial = via plugin or limited · — = not available

Simple pricing

Start with a 14-day Pro trial. Annual plans save 40%.

Basic

$20/mo

or $144/yr (40% off) · 14-day free trial

Core editing with a monthly allowance of AI assistance.

Most popular

Pro

$40/mo

or $288/yr (40% off) · 14-day free trial

10× Basic AI allowance and Deep Research with verified citations.

Prices in USD. Tax/VAT added at checkout based on your location. Payments processed by Paddle.com (Merchant of Record).

Try it on your next paper.

Start with the 14-day Pro trial. No credit card. Cancel any time.

Create your account