Making a PDF is byte-writing, not magic, which is why your browser can do it without sending anything anywhere: take images, text, Markdown, or a table, lay them onto pages, write the bytes, hand you the download. The craft is in the layout decisions, page size, margins, image fit, and in a few traps like text encoding. This guide covers building PDFs from each input type, starting with the workhorse, our free image to PDF converter.
In this guide
Pages, points, and why A4 is 595 wide
PDF measures pages in points, 72 to the inch, a unit inherited from typesetting. That makes the two standard pages:
- A4: 210 × 297 mm = 595.28 × 841.89 points
- US Letter: 8.5 × 11 inches = 612 × 792 points
Every PDF generator is positioning content on one of these grids (or a custom one), and most layout questions reduce to points arithmetic: a 1-inch margin is 72 points, usable A4 width inside it is about 451 points, and so on. You never need to compute this by hand to use the tools, but knowing the grid exists explains their options: page size, orientation, and margins are the whole coordinate system being chosen.
Images to PDF: the scanner replacement
The most common creation job: phone photos of documents, receipts, or whiteboards into one orderly file. Three decisions shape the result:
- Fit mode. “Fit to page” letterboxes each image inside the printable area with no cropping; “fill” covers the page and crops edges. For documents, fit; for photo albums, taste.
- Orientation per image. A good tool rotates the page to match each image rather than forcing landscape photos sideways onto portrait pages. Mixed orientations in one PDF are perfectly legal, as the merge guide notes.
- Order before, not after. Arranging images in the tool costs seconds; reordering pages in the produced PDF is a second job. The converter lets you sort before generating.
One expectation worth setting: the PDF will weigh roughly what the images weigh, since they go in as-is. Twenty 3 MB photos make a 60 MB PDF, and the right fix is resizing photos before conversion, the same source-first logic as the compression guide.
Text to PDF: the timeless format
Plain text into a paginated, printable, everywhere-openable document: notes, logs, code, license files. The decisions are font (monospaced keeps code and ASCII tables aligned; proportional reads better for prose), size, and margins, and the tool wraps and paginates from there. The one real trap is encoding: text with accents, Greek, or any non-Latin script must be handled as UTF-8 with a font that carries those glyphs, or the output shows the familiar mojibake garbage. Our text to PDF tool treats input as UTF-8 for exactly that reason; if you have ever opened a “final” PDF full of question marks where the accents were, encoding was the culprit.
Markdown to PDF: structure for free
Markdown carries structure that plain text lacks, headings, lists, emphasis, code blocks, tables, so converting it yields a formatted document with no word processor involved: README files, meeting notes, documentation. The Markdown to PDF tool renders the structure to styled output, headings sized, lists indented, code in monospace. The practical limits are Markdown’s own: page-level design (columns, precise placement, branding) is not what the format describes, so when a document needs design rather than structure, Markdown to PDF is the draft step, not the final one.
Tables and Base64: the special cases
- CSV to PDF table. Spreadsheet data into a paginated, printable table, with the CSV to PDF tool handling column widths and page breaks. Wide tables fit better in landscape orientation, and very wide ones honestly belong in a spreadsheet, not on paper.
- Base64 to PDF. APIs, email systems, and databases often hand over PDFs as Base64 text, a long block of letters and digits encoding the file’s bytes. The Base64 to PDF tool decodes that block back into the actual file, locally, which beats pasting a possibly-sensitive document blob into a random decoding site.
All of the creation tools share the property that makes the whole family work in a tab: writing a PDF needs no server, just bytes in the right order, which is the theme of our PDF pillar.
Frequently asked questions
Will the PDF look the same on every device?
Yes, that is PDF’s founding promise: the layout is fixed at creation, with fonts and positions baked in. It is why forms, invoices, and anything signed travel as PDF rather than as editable documents.
Is the text in my generated PDF selectable?
From text, Markdown, and CSV: yes, real selectable, searchable text. From images: no, an image page is pixels, and making it searchable is the OCR problem described in the extraction guide.
What resolution should photos be for a clean PDF?
For screen reading, anything from a modern phone is more than enough. For print, aim for roughly 150 to 300 dpi at the printed size, which a standard phone photo of a document comfortably exceeds. More resolution past that adds weight, not visible quality.
Can I combine the outputs, like text pages plus image pages?
Yes: generate each part with its tool, then combine with merge PDF. Build the parts, then bind the book.