Your paper lives across too many places.
Notes, LaTeX files, PDFs, figures, references, code, and advisor comments drift apart unless the workspace keeps them connected.
SuperWriterA desktop workspace for planning, writing, citing, revising, and preparing scientific papers with the full project context in view.
Purpose-built Skills connect literature work, drafting, citations, code-to-methods, visual briefs, review, and response letters inside the same local project.
Project
files, LaTeX, PDF, figures, versions
15
research-writing Skills
MCP
Zotero, citations, files, web, visuals
From project intake to revision response, SuperWriter keeps files, evidence, AI edits, and submission artifacts aligned.
Students and early-career researchers need structure, context, and cost clarity. SuperWriter keeps the whole writing system in one local workspace.
Notes, LaTeX files, PDFs, figures, references, code, and advisor comments drift apart unless the workspace keeps them connected.
A chat box can polish a paragraph, but it rarely tracks evidence, figures, citations, versions, and submission artifacts together.
Starter routes AI through SuperWriter credits, so users can begin without registering model providers or configuring base URLs.
SuperWriter shows an estimate before a task starts and settles after real usage, with balance and usage visible in the app.
Move from intake and storyline to drafting, figures, review, submission package, and revision response without losing context.
Natural-language routing sends paper, review, literature, grant, patent, defense, proof, promotion, and audit work to the right Skill.
Starter and Pro can use built-in metered AI. Tasks are estimated before running and charged after actual input and output usage.
Import local or GitHub repositories and generate a paradigm-aware Methods Card, citation candidates, and availability stubs.
Connect Zotero, AI4Scholar, PubMed, arXiv, Semantic Scholar-style discovery, full-text reading, and BibTeX workflows through MCP.
Keep project files on your machine with LaTeX editing, live PDF preview, local history, diff review, and explicit AI-assisted changes.
Prepare graphical abstract, TOC, mechanism, and method-flow briefs with prompt records and generated visual artifacts when configured.
Pro users can connect their own official or OpenAI-compatible API route and bypass SuperWriter credit spending for those calls.
The v2 Skill system is organized around real research deliverables, not isolated prompts.
Credits are the AI usage balance inside SuperWriter. Non-AI desktop features remain available without spending credits.
Task cost examples
| Task | Economy model | Flagship model |
|---|---|---|
| Paragraph polish or abstract rewrite | a few credits | tens |
| Figure captions or section-level drafting | tens | hundreds |
| Code repository to Methods Card | tens to hundreds | hundreds |
| Long literature review or NSFC chapter pass | hundreds | thousands |
If bundled credits run out, credit top-ups are available later. They are not required to use the non-AI desktop workspace.
SuperWriter keeps the desktop project as the base layer, then connects specialized tools when the task needs them.
SuperWriter starts from the reality of academic writing: the hard part is keeping claims, evidence, figures, citations, and revisions aligned. The assistant works inside the same local project as your files.
Built by PolyAI with strong support for materials, chemistry, polymers, AI-for-science, and Chinese academic workflows.
Starter removes the API setup hurdle for individual users. Pro preserves the BYO path for teams that already operate their own model budget or relay.
Local
project files
MCP
research tools
Credits
usage ledger
Both plans include the complete SuperWriter desktop app. The difference is bundled credits and whether you can bring your own API.
Best for individual researchers
5,000 bundled credits
Built-in SuperWriter AI gateway
For heavy users and labs
30,000 bundled credits
BYO API for official or compatible providers
Download SuperWriter, buy a Starter or Pro activation code, and keep your next paper's files, AI work, and revisions in one desktop project.