Research memorythat actuallyremembers.
Upload your papers, log experiments, chat with your knowledge. Draft publication-ready manuscripts with sentence-level citations that are actually verified.
A system that remembers everything you read, reasons over it with citations you can verify, and writes manuscripts grounded in your own data.
Your entire research library, searchable in seconds.
Drop PDFs, CSVs, images, notes. SuperLabCoat reads every file, pulls out what matters, and links it into a knowledge graph. No manual tagging, no folder hierarchies, no lost files.
- Auto-extractionTables, figures, abstracts, DOIs pulled from every upload.
- Search by meaningAsk in plain English, not keywords. Results in under 200ms.
- Multi-modalPDF · images · spreadsheets · markdown · plain text.
Ask anything. Get cited answers.
The copilot searches your workspace, finds the most relevant passages, writes answers grounded in those source paragraphs, and verifies each citation against the cited paper's content, not just a link.
- Grounded in your libraryOnly cites papers you've actually uploaded and connected.
- Streaming stepsWatch it search, rank, compose, and verify. Transparently.
- No fake referencesEvery [N] is checked against the source paper before it's shown to you.
Every paper, experiment, connected.
A living knowledge graph builds itself as you work. Papers cite papers. Experiments test hypotheses. Results support or contradict them. See your reasoning structure, not just your files.
- Rich relationshipscites · tests · produces · supports · contradicts · derives-from · references · uses · related-to
- 2D and 3D viewsNavigate by meaning. Spot contradictions before reviewers do.
- Hypothesis trackingWhich results confirm it? Which contradict? One glance.
From data to publication-ready figures.
Ask for a chart, a diagram, a stats test, a slide deck. The system generates it, returns the file, and saves it to your artifacts library. No notebooks. No copy-paste. No separate analysis tool.
- 10 built-in toolscharts · PPTX · DOCX · XLSX · PDF · t-test · ANOVA · networks · symbolic math · data analysis
- Isolated executionAnalysis runs in a hardened sandbox with no outbound network. Your data stays in your workspace.
- Every output savedArtifacts are versioned, searchable, reusable. Not throwaway chat responses.
A full manuscript draft, with every citation verified.
Point the draft tool at your workspace. It gathers your experiments, pulls in the papers you've actually cited, writes abstract · intro · methods · results · discussion, embeds tables and figures, and hands you a Word document ready for submission.
- Section-by-section draftingEach section is written with the sources that matter most to it.
- Auto tables + figuresResults tables are built from your own data. IC₅₀ and fold-change charts generated on demand.
- Citation reportEvery [N] reference audited against the cited paper's text, with confidence scores.
Doxorubicin resistance in MCF-7 breast cancer cells is primarily mediated by ATP-binding cassette transporters [1]. We screened four ABCB1 inhibitors in combination with doxorubicin and identified cyclosporine-A as the most synergistic combination (Combination Index = 0.28)[2]. IC₅₀ values fell from 12.4 μM to 1.2 μM (p < 0.001, n=6)[3].
Your research stays yours.
We built SuperLabCoat because we were frustrated with tools that vacuumed up our workspaces to train public models. Your data is never used for training. Period.
Workspace isolation
Each workspace is a hard data boundary. Your uploads never cross into another lab's context.
Sandboxed execution
Generated code runs in an isolated sandbox with no outbound network. Your data never leaves your workspace.
Encrypted at rest
All uploads encrypted in enterprise-grade cloud storage. Your database is access-controlled and continuously backed up.
SSO + audit logs
SSO on Lab plans. Every action logged. See exactly who did what, when.
Built for the people who actually do the work.
Stop losing track of what you read. A knowledge graph that catches the contradictions you'd otherwise miss until peer review.
Turn a week of cross-referencing into an afternoon. Draft methods and results sections from the work you've already logged, with verified citations.
Every citation audited against its source. No hallucinated references. No embarrassing retractions. Just answers grounded in your own library.
Everything isfreeuntil June 30.
Every feature, no limits, no card needed. We're heads-down listening to early users. Plans and pricing land after the launch window.
- Unlimited papers, experiments, hypotheses
- Citation-grounded copilot
- Knowledge graph · 2D and 3D
- All sandbox tools — charts, stats, slides, drafts
- Publication draft with verified citations
- Sandbox isolation, encrypted storage, your data stays yours
Answers before you ask.
SuperLabCoat builds a real knowledge graph out of your uploads: papers, experiments, hypotheses, results, all linked to each other. When the copilot answers, it's not just pattern-matching your files; it's reasoning over the structure of your research. Every citation is verified against the source paper before it's shown to you. And the draft tool produces Word documents with tables and figures generated from your own data, not templated filler.
No. Never. Your workspace is isolated and your uploads are never used to train any AI model, ours or anyone else's. Your data is yours. Period.
Yes. The Researcher plan includes 1,000 papers. For larger libraries the Lab plan is unlimited. Search stays fast (under 200ms) even at thousands of papers.
PDF (with automatic table and figure extraction), CSV, Excel, TSV, TIFF, PNG, JPG, Markdown, and plain text. DOIs are detected automatically and bibliographic metadata is pulled in for you.
Yes, on the Lab plan. Shared workspaces with comments, an activity feed, per-item attribution, and an admin dashboard that shows who's contributing what.
It pulls together the experiments, results, hypotheses, and protocols you've logged in your workspace, scopes the references to the papers you've actually cited, and writes the manuscript section by section: abstract, introduction, methods, results, discussion. Each section is grounded in the sources that matter most to it. Results tables are built from your own data. Figures (IC₅₀, fold-change, etc.) are generated fresh. Every citation is audited against the source paper and shipped with a confidence report alongside the final Word document.
Stop re-readingthe same papers.Start publishing.
Free forever plan. Your first publication draft in under 20 minutes.