Export Emails as Markdown or PDF — and Why Format Matters More Than Ever

EmailShot has always been about getting an email out of your inbox and into the world — a shareable link anyone can open in a browser, no Gmail account required. Today we are adding two new export formats: Markdown and PDF, available right from the share button on any email page.

On the surface, these are simple convenience features. But the timing is interesting. As more people build personal knowledge bases powered by LLMs, the format in which you capture information has started to matter a great deal. Let's look at what's new and why it's more useful than it might first appear.

What's New

Open any EmailShot page — emailshot.io/p/your-email — and click the dropdown arrow next to the Share this email button. You'll now see two new options:

• Get Markdown: opens a plain-text Markdown version of the email in a new tab. The subject becomes an H1 heading, the From/To/Date headers are preserved, the HTML body is converted to clean Markdown (links, bold, lists, headings all intact), and there's a link back to the original at the bottom. You can save it directly or copy it into any editor.
• Save as PDF: opens a print-ready version of the email in a new tab and triggers the browser's print dialog. The PDF includes the original email HTML with proper formatting, plus a QR code at the bottom linking back to the live EmailShot URL — so anyone who encounters the printout can scan their way back to the original.

Both are available without any login or extra configuration.

Your Inbox Is Full of Source Material

Here's the thing about email: it's one of the richest and most underrated sources of knowledge most people have. Newsletters from experts you trust. Research your colleagues shared. Client feedback that shaped a product decision. Industry analysis forwarded by a friend.

All of it sits in your inbox, mixed in with receipts and meeting invites, and most of it never makes it anywhere more permanent.

Andrej Karpathy recently published a concise pattern for thinking about this problem: LLM Wiki. The core idea is that instead of retrieving raw documents at query time (the standard RAG approach), you maintain a persistent, curated wiki of Markdown files — one that an LLM incrementally builds and maintains as you feed it new sources. When you add a document, the LLM doesn't just index it; it reads it, extracts what matters, and integrates the findings into the existing wiki — updating entity pages, noting contradictions, strengthening synthesis.

The pattern has three layers:

• Raw sources — the original documents, never modified
• The wiki — LLM-generated Markdown pages: summaries, entity pages, syntheses
• The schema — instructions (a CLAUDE.md or AGENTS.md file) that tell the LLM how to maintain the wiki

What makes this work, Karpathy argues, is that the wiki is a compounding artifact. Cross-references already exist. Synthesis reflects everything consumed. It gets richer with every source you add. Humans curate and direct; LLMs do the bookkeeping.

Email is exactly the kind of material this system thrives on — and until now, getting an email into that system meant copy-pasting, forwarding to yourself, or some janky browser extension.

Markdown Export for LLM Workflows

Markdown is the lingua franca of LLM knowledge bases. Obsidian stores everything as Markdown files. Claude Code's project memory is Markdown. Most note-taking tools that connect to LLMs expect Markdown. So when EmailShot gives you a Markdown export, it's giving you a file you can drop directly into any of these systems with no conversion needed.

A few concrete workflows this enables:

Building an email archive in Obsidian. You subscribe to half a dozen newsletters that consistently contain insights worth keeping. Instead of letting them pile up as read/unread in Gmail, you EmailShot them and drop the Markdown exports into your vault. The LLM gets to read the actual content — with preserved links, formatting, and metadata — rather than a screenshot or a stripped text dump.

Feeding source material to an LLM agent. If you're running the LLM Wiki pattern, your workflow for ingesting a new newsletter is: EmailShot it, download the Markdown, drop it in the raw sources folder, ask your agent to ingest it. The agent reads the full email, extracts key concepts, and updates relevant wiki pages. The Markdown link at the bottom points back to the original EmailShot URL, so your wiki entries can cite the source.

LLM-assisted summarization. Even without a full wiki setup, paste a Markdown-exported email directly into Claude or ChatGPT and ask: "What are the three most important things in this email, and what should I do about them?" The structured format — clear metadata block, clean body, no HTML noise — gives the model less to filter through and more to work with.

Team knowledge bases. Karpathy mentions this pattern for business wikis fed by Slack, transcripts, and documents. Email belongs in that list. A marketing team ingesting competitor newsletters, a product team capturing customer feedback threads, a legal team archiving contract communications — the Markdown export makes each of these a one-click operation.

PDF Export for Everything Else

Not everything is an LLM workflow. Sometimes you need to hand someone a piece of paper, file something in a document management system, or give a client a record they can keep without depending on a URL staying alive.

The PDF export handles the opposite problem from Markdown: instead of stripping structure down to text, it preserves the visual richness of the original HTML email — layout, images, typography — in a format anyone can open without special software.

The QR code at the bottom is there for a specific reason. Printed documents lose their connection to the original. A month after someone receives your printout, they'll have no idea where the email came from or whether there's a live version somewhere. The QR code keeps that thread intact — scan it and you're back at the EmailShot URL with the original email, all attachments accessible, shareable again.

This makes it particularly useful for:

• Client deliverables: print a campaign email you designed for a client, QR code links back to the live version for easy sharing
• Compliance and audit trails: a printed email with a QR back to the permanent EmailShot URL gives you both a physical record and a digital reference
• Presentations: print a customer testimonial or key approval email; the QR lets anyone in the room verify and share the original
• Filing: some teams still maintain physical files alongside digital records; the QR bridges the two

Format Is Infrastructure

The broader point is that format determines where information can go. A great newsletter read and forgotten in Gmail has near-zero knowledge value. The same newsletter as a clean Markdown file in Obsidian, cited by five wiki pages, cross-referenced with similar content, and queryable by an LLM — that's a different thing entirely.

Karpathy's observation is that humans abandon wikis because the maintenance is tedious. LLMs don't abandon them — but they need the source material in a form they can actually work with. Markdown export from EmailShot is a small piece of that infrastructure.

PDF is the other direction: taking something ephemeral and giving it permanence and portability, while keeping the digital thread alive through the QR code.

Both formats are available now on any EmailShot page. If you haven't created a shareable email link yet, the getting started guide walks you through it in under a minute. And if you're building a personal knowledge base or experimenting with the LLM Wiki pattern, the Markdown export is a good place to start.