PDF
Your PDF
Paste your PDF. Nothing to upload.
←PDF
Make a research report
Turn a folder of PDFs into a research report
Upload several. One question, what the documents agree on, and exactly where they contradict each other.
Free to try, no card. Your PDF opens in Ancher.
Document structure in orderPage numbersTables as tables
Kept in parts
Every one of these stays separately addressable instead of collapsing into one block of text.
The questionKey findingsEvidenceCounterpoints
Research report
A structured answer to a question, with the evidence and the counter-evidence attached.
The unit of work is the collection, not the file. Ancher groups claims across every PDF you add by the question they answer, compares methods side by side from the tables, and names the contradictions rather than averaging them — because a synthesis that resolves every disagreement has deleted the information you needed most.
The honest version
A synthesis that agrees with itself everywhere is a summary of your search query
Assemble a set of documents that all say the same thing and any tool will produce a confident, coherent report. That report tells you nothing, because the coherence came from the selection rather than from the field. A useful synthesis names its disagreements and shows which findings rest on a single source.
One page of one PDF, at p.9, read two ways.Illustrative. The figure is invented; the gap it falls into is not.
Paper A concludes the effect is robust. Paper C, same measure, reports
Table 2, page 9, Paper C — not readThe specific never reaches the draft
Research report
Table 2, page 9, Paper C
no significant effect
from p.9
Lands in the draft, with its timecode
- Contradictions between documents are named on both sides with page citations, not smoothed into a middle position.
- Each finding shows how many documents support it, so single-source claims cannot pass as consensus.
- Method details come from the tables, which is what makes a real comparison possible rather than a prose gloss.
What comes out
Research report, section by section
A structured answer to a question, with the evidence and the counter-evidence attached.
The question
One question the set actually addresses, plus how many documents, what date range, and how many are preprints.
Key findings
Claims grouped by question, each showing how many documents support it and which.
Evidence
The figures behind each finding, from table cells, with document and page.
Counterpoints
Direct contradictions, named on both sides — the section that makes the report worth more than any one paper.
What to check next
Outcomes and populations nothing in the set measures, plus the works everyone cites.
Multi-PDF research report prompt
The link on its own gets a summary. This is the instruction that produces the 5 sections above, in that order, with the rules that keep them honest. Paste it with your PDF — in Ancher, or in whatever assistant you already use.
Across every PDF in this collection, produce a research report. (1) State the question, how many documents, the date range, and how many are preprints. (2) Group claims by question; for each finding say how many documents support it and name them. (3) Give figures from methods and results tables with document and page. (4) List every direct contradiction, naming both sides — do not resolve them. (5) Identify outcomes or populations nothing in the set measures. Never state a claim without a source and page.
The detail, if you want it
How each part of a PDF becomes a section
| From the PDF | Into | Why |
|---|---|---|
| Document structure in order | Key findings | Section headings across documents cluster into themes more reliably than full-text similarity, because authors name their own contributions. |
| Tables as tables | Evidence | Sample size, design and effect size live in the methods tables, which is what makes a genuine comparison possible. |
| Page numbers | Counterpoints | A contradiction is only useful if both sides are checkable, which needs a page on each. |
| Document structure in order | What to check next | Reference sections across documents are comparable once they survive as elements, and works cited by everything in the set are what your scope has to account for. |
| Word and PowerPoint too | The question | Because decks and Word files take the same path, a working paper and its presentation can sit in one synthesis. |
Everything Ancher keeps from a PDF
Document structure in order
Headings, body and captions arrive as separate elements, so the document's own outline survives ingest.
Page numbers
A page-level anchor is what makes a citation checkable by someone else holding the same file.
Tables as tables
Tables come back as their own elements rather than as flattened lines, so a figure stays attached to the row and column that give it meaning.
Embedded images and figures
Figure captions frequently carry the tightest statement of a finding in the whole document.
Reading order
Elements come back in sequence, which is what lets a section be reassembled rather than guessed at.
Word and PowerPoint too
The same path takes .docx and .pptx, so a deck and its write-up can sit in one collection.
The four steps
- Add the PDF. Upload the file. Ancher parses it into ordered elements rather than one flat string, keeps the page number on each, and pulls out the embedded images.
- Nothing gets flattened. Document structure in order, page numbers, tables as tables, embedded images and figures, reading order, word and powerpoint too stay separately addressable instead of collapsing into one block of text.
- Ask for a research report. Ancher fills the question, key findings, evidence, counterpoints, what to check next using the mapping above, not a generic summary pass.
- Check it before you use it. Every specific keeps a link back to where it came from in the PDF, so a claim can be verified without reopening the source.
Where this goes wrong
- A set that agrees with itself is a search-query artefact. Deliberately add the strongest dissenting document you can find before drawing any conclusion.
- Preprints and peer-reviewed work carry different weight. If review status is not tracked per document, the synthesis quietly overstates its own confidence.
- Below about eight documents, reading them yourself is faster and you will retain more. The value here comes from cross-document structure, which needs enough documents to be real.
Questions people ask
Will it invent citations?
Every claim binds to a document and a page inside your own collection, so a claim with no source has nothing to attach to. That is the specific reason to synthesize from files you added rather than from open-web recall, where fabricated citations are the standard failure.
Can it tell me what the field is missing?
It can show what your set is missing — outcomes nobody measured, populations nobody sampled. Whether that is a real gap in the field or a gap in your search is a judgement that stays with you, and the report is written to keep that distinction visible.
Can I mix PDFs with articles and videos?
Yes. A claim supported by three papers and two conference talks is a different kind of finding from one supported by papers alone, and the report can only make that distinction if both are in the collection.
Bring your own PDFs
Everything you save lives in one workspace, so the research report is built from your sources — not from a model's memory of the internet.
Open Ancher →Not a PDF? The same research report also comes from YouTube video, X thread, LinkedIn post.