To chat with multiple PDFs reliably, ask questions that name what you want compared, require every claim to say which document it came from and quote it, and look at which documents the answer actually read. The common failure is not a wrong sentence — it is a source that was never looked at, and an answer that sounds complete anyway.

One PDF is the easy case. The questions worth asking usually span ten: which of these papers use the method, where do these reports disagree, what does the contract say that the summary left out.

AI invents citations because, to a language model, a reference is just more text that should look plausible — and a plausible title, author and page number are easy to produce without any document behind them. The check that catches most of it is simple: require the exact quoted words, and match them against the source text before trusting the claim.

That is a mechanical check, so software can do it. Most tools do not, and the gap is where the embarrassing stories come from.

An AI mind map is worth using again because the old weakness was never the diagram — it was that a blank map only holds the angles you already thought of. When each branch can apply a named thinking model to your actual material, the map stops reflecting your habits and starts widening them.

Mind mapping has been quietly out of fashion for a decade, and the usual explanation — that the tools were clunky — has never been true. The tools were fine. Something else was wrong, and it is worth naming precisely, because it turns out to be the part that changed.

To read a research paper with AI without being misled, ask for the paper's central claim and what it rests on before you read anything, insist that every point comes with a verbatim quote and a page, check the two quotes that matter most, and only then read those sections properly yourself. The AI finds the way in; you still do the believing.

The usual approach — paste the PDF into a chat window and ask for a summary — is fast and feels productive. It also gives you a fluent paragraph you cannot evaluate without reading the paper, which is the thing you were trying to avoid.

MindMax is a desktop app for macOS and Windows that reads, thinks and writes with you inside a folder you already own. Answers about your documents carry verbatim quotes you can click back to the page, and every AI edit to your writing arrives as a tracked change you accept or reject.

There are a lot of tools now that will read a PDF and talk to you about it. We built a different shape of thing, and it is worth being clear about why.

The first public beta build.

What is in it​

• Reading. PDFs, EPUBs, Word, PowerPoint, Excel and Markdown open from a folder you choose. Annotations anchor to the real position in the document and are stored alongside it.
• Region questions. Drag a box around a chart or table and ask about just that part of the page — useful when a paper stops making sense at a figure rather than at a paragraph.
• Checkable answers. Questions about a document come back with verbatim quotes that link to the page they came from. Quotes that do not match the source are dropped rather than shown.
• Canvas. 37 built-in AI Skills across six groups, for taking one question and working it over from several directions.
• Long-form writing. Outline approval before drafting, an argument ledger across chapters, and revisions presented as marked-up changes.

Known rough edges​

• Builds are not signed yet, so both macOS and Windows will warn on first open. The installation guide walks through it.
• Multi-column PDFs with interleaved figures are still the hardest extraction case; quotes from those pages are worth clicking through.
• No mobile app, no web version. Not planned.

Where to start​

Understand a paper in about 30 minutes is a good place to find out whether this is for you.