Write a thesis you can defend.

CiteDash AI is an AI research operating system for graduate students and academics. It carries a thesis through the whole lifecycle (research question, literature search, PDF reading, evidence synthesis, data analysis, drafting, revision, citations, and compiling to DOCX/PDF) in one workspace where every citation resolves to a real paper.

General AI chatbots routinely invent citations: plausible-looking references to papers that don't exist. In CiteDash AI that is impossible by construction: every citation is a database object that must resolve to a real paper, and every cited claim is checked against the actual full text of its source by the built-in Fact Checker before you ever see it. Claims that can't be verified are flagged, never silently shown.

No. There are no free-text references anywhere in the product. A citation that does not resolve to a real paper is rejected before it reaches your draft, and claim-to-source verification runs against the paper's embedded full text, not its abstract, and never the model's memory.

PhD and Master's students writing a thesis or dissertation, researchers writing papers and systematic reviews, and supervisors who need to trust what their students submit. If you have to defend every claim you write, CiteDash AI was built for you.

Any citation-driven field. The live search spans OpenAlex, PubMed, Semantic Scholar, and arXiv, so the sciences, engineering, medicine, social sciences, and the humanities are all covered. Verification, retraction blocking, citation styles (APA, MLA, Chicago, IEEE), and DOCX/PDF/LaTeX compiling are discipline-agnostic. If your work cites papers, CiteDash AI fits.

CiteDash AI has three plans: Starter at $10/month (1,000 credits/month), Pro at $20/month (2,500 credits/month), and Ultra at $99/month (15,000 credits/month). Yearly billing is 20% off ($96.00, $192.00, and $950.00 per year) with the whole year's credits granted upfront. Credits are the usage currency: they spend on AI drafting, fact-checks, searches, and compiles across all your projects. You can preview CiteDash in demo mode without signing up before you subscribe.

Yes. Every account gets a personal referral link on the Settings page. When someone signs up through your link and creates their first project, you earn 501 bonus credits, added to your balance automatically. The reward is always shown current on the pricing page and your Settings card. Each new user counts once, and self-referrals never count.

Credits pay for the actions that cost real compute: roughly 1 credit fact-checks a claim against its source, 1 credit runs a live literature search across four sources, about 10–30 credits AI-drafts a chapter section with verification, and 5 credits compile the full thesis to DOCX/PDF/LaTeX. Your plan grants credits every month and they work in every project on your account. Cheap discovery (corpus browsing, reading, your library) stays unlimited on every plan.

Nothing is lost. Your projects, library, drafts, and citations all stay exactly where they are. You keep full read access. Only the credit-metered actions (AI drafting, fact-checks, live searches, compiles) pause until your plan refreshes at the start of the next billing month, or you move to a higher plan for more monthly credits. You'll see a low-balance notice before you hit zero, never a surprise.

Yes. Plans are month-to-month (or yearly if you choose) and you can cancel or change plans anytime from your account. There's no lock-in. You keep access through the end of the period you've paid for, and your projects, library, and compiled documents remain yours to export.

Every search reaches your own indexed corpus plus the live wider literature (OpenAlex, PubMed, Semantic Scholar, and arXiv) reranked into one list. Open-access PDFs are fetched automatically so papers become fully readable and citable, and you can upload any PDF yourself.

You remain the author: nothing enters your thesis without your review. Every AI-drafted claim carries a verified citation with sentence-level provenance back to the exact source passage, retracted papers are blocked at every step, an originality screen runs on every compile, and a complete audit trail plus an AI-use disclosure document is generated for your examiner or supervisor.

CiteDash AI is built to be used openly, not hidden. You choose the sources, direct every draft, and review everything before it enters your thesis: the AI drafts only from papers in your own library and you stay the author throughout. Every AI-assisted step is logged into an audit trail and an AI-use disclosure document you can hand to your committee, the kind of transparency many university AI policies center on. Policies differ, so check your institution's rules and your supervisor's expectations, and use the disclosure to be upfront about how you worked.

CiteDash AI runs a verbatim-overlap originality screen on every compile so accidental overlap is caught before submission, and the honest path with AI detectors is disclosure rather than evasion: the generated AI-use disclosure documents exactly how AI assisted your work. AI detectors are known to be unreliable in both directions, and no tool can honestly promise you a particular detector score. What CiteDash gives you instead is the stronger position: verified sources, sentence-level provenance for every claim, and a complete audit trail showing the work is yours.

Yes. Import your Zotero or Mendeley library as .bib (your citation keys are preserved and round-trip back out), import existing .docx chapters (their citations relink to real papers, and a retracted source is flagged immediately), and upload any PDFs you've collected. From there, everything you draft or redraft is verified claim by claim as you write. Your work also leaves with you anytime, as DOCX, PDF, LaTeX, and BibTeX exports.

AI research assistants are strong at retrieval: they search real literature and summarize real papers. Retrieving a real paper is not the same as verifying a claim, though: those tools still leave you to check that what the AI wrote is actually what the paper says. CiteDash AI checks the sentence you wrote against the source's full text before you ever see it, drafts chapters grounded in your own library rather than one-off summaries, and carries the work through to a compiled, examiner-ready document with an audit trail, retraction blocking, and an AI-use disclosure. Discovery tools end where your thesis begins.

You do. Your uploads, projects, drafts, and library belong to you and are isolated to your account. Other users never see them, and you can export or delete your work at any time. Your documents are used to power your own research workspace, not to train public models. Uploaded PDFs are stored so your papers stay readable and citable, and can be removed on request.

Yes. The Thesis Assembler compiles your chapters, front matter, and references into DOCX, PDF, and LaTeX, with your citation style (APA, MLA, Chicago, or IEEE) applied consistently, and your BibTeX library round-trips with your own citation keys intact.

Yes. The guided Thesis → Manuscript flow converts a finished thesis into a journal-ready manuscript in one sitting. Import your thesis .docx (its citations relink to real papers) or pick your CiteDash chapters, choose a target journal by real semantic fit, condense chapter by chapter with drafting grounded in your library's full text, re-verify every claim through the compression, and build the submission package: cover letter and AI-use disclosure, plus the compiled manuscript PDF once you've assembled the document in the Thesis Assembler.