What is RTF, and what changes when it becomes a PDF?
Rich Text Format is a document format made of plain text. Open an .rtf in a text editor and you can read it: the words are there, with instructions around them written as backslash codes — \b for bold on, \i for italic, \fs36 for an 18 pt run, \qc for a centred paragraph, \par for the end of one. Microsoft published it in 1987 so that two word processors that agreed on nothing else could still hand each other a formatted document, and it worked well enough that thirty-nine years later it is still what a case-management system, a transcription service or an old accounting package emits when you ask it for something with bold in it.
RTF's strength is also its problem. Being text, it survives email and old backups intact. Being a word-processor format, it needs a word processor. Nothing on a phone opens it nicely, a client who receives one is as likely to get a wall of backslashes as a document, and it reflows — the same file breaks its lines differently on two machines, which is exactly wrong for anything anyone has to sign.
A PDF is the other kind of document: a fixed description of marks on a page. This character, this font, this size, at these coordinates, on a page this big. It does not reflow, it opens everywhere, and it prints the same in Lisbon as it did in Lagos. So converting RTF to PDF is one job, and it is a layout job — take styled text with no page in mind and decide where every word lands on paper, without losing the formatting someone opened a word processor to add. That is what the RTF to PDF converter does, in your browser, with the file staying on your machine.
How the RTF to PDF conversion works
Two passes. The first reads the document, the second draws it.
The parse walks the RTF character by character, keeping state the way the format expects: braces push and pop, so a {\i up} turns italic on and back off at the closing brace without touching the rest of the sentence. Sizes are the interesting part. RTF measures type in half-points, and its default body size is 24 half-points — 12 pt. A run is not stored as "18 pt"; it is stored relative to the body, and that relationship is what the converter keeps.
rendered size = base size × (half-points ÷ 24) — a heading written as \fs36 is 36 ÷ 24 = 1.5 × body text. At the default 12 pt base it draws at 18 pt. Set the base to 16 and the same heading draws at 24 pt. The document keeps its own proportions instead of being flattened to one size.
The second pass is layout, and it happens before you download anything. Every word is measured with the real advance width of each of its glyphs in the face the PDF will use, so the tool knows "Agreement" in 18 pt Times bold is almost exactly 30 mm wide and where the line has to end. Wrap, page break, page count — all decided up front. That's why the button reads Make the PDF (7 pages) before you press it, not after.
Then the characters go into the PDF as characters, through its text operator — not painted onto a canvas and pasted in as a picture, which is the cheap way to build a converter and the way that produces a file nobody can search or copy out of.
What survives the conversion, and what doesn't
Every converter loses something. The useful question is which things, and whether you find out before or after you send the file. Here is the list.
| What's in your .rtf | In the PDF | Notes |
|---|---|---|
| Bold, italic, underline | Kept | Including all three at once, and mid-sentence changes |
| Heading and body sizes | Kept, relative | A heading 2× the body stays 2× the body |
| Left, centre, right, justified | Kept | Justified lines fill both margins, except a paragraph's last |
| Left and first-line indents | Kept | Measured in twips: 720 twips = 0.5 inch = 12.7 mm |
| Tabs | Kept as real stops | Every half inch, not a run of spaces |
| Explicit page breaks | Kept | Counted and reported back to you |
| Accents, curly quotes, dashes | Kept | café, naïve, señor, em dashes, ellipses, €, ™ all draw |
| Table cell text | Kept, flattened | Cells laid out on tab stops; grid lines not drawn |
| Images | Dropped | Counted and named before you build the file |
| Text colour and highlighting | Dropped | Everything is drawn in black |
| Headers, footers, footnotes, comments | Dropped | Skipped deliberately, not half-drawn into the body |
| Chinese, Cyrillic, Greek, emoji | Shown as "?" | Counted by code point and listed before download |
The last four rows are the honest part. A text-only pipeline drawing in a reader's built-in fonts cannot draw a photograph, and no standard PDF font holds a single Chinese glyph — so the choice isn't between losing them and keeping them, it's between telling you and not telling you. This page counts and names them while you're still looking at the settings: 1 character can't be drawn by a PDF's built-in fonts and will show as "?": 中 (U+4E2D). Most converters let you discover that by opening the download.
Most converters are also three clicks deep inside a document platform that wants a per-seat subscription, a fourteen-day trial and your work email before it will look at a 40 KB file you already own. The RTF to PDF converter is just the converter — the full tool, not a teaser for a paid one. There's no Pro tier to graduate to, and 10% of every dollar Microapp earns goes to charity off the top, audited quarterly.
An RTF to PDF conversion, start to finish
Take a short contract: a centred 18 pt bold heading, a body paragraph with a first-line indent, an italic phrase, an underlined word, and two tabbed lines at the bottom. Saved as .rtf, its markup reads:
\qc \fs36 \b Service Agreement \b0 \fs24 \par \ql \fi720 The fee is \i payable monthly \i0 , due on the \ul first \ulnone of each month…
Drop that file in and the numbers appear immediately: 1 page, 28 words, 4 paragraphs, 154 characters, under a line reading Body text at 12pt Serif (Times), 20mm margins. 3 runs of bold, italic or underline kept. A note adds that the document is set in Calibri and will be drawn in the face you pick. Here is where each piece lands on an A4 page at the defaults.
The heading. \fs36 is 18 pt against a 12 pt base, so it draws in Times bold at 18 pt — and because \qc centres it, the word "Service" starts 79.35 mm across a 210 mm page.
The indent. \fi720 is 720 twips, and a twip is a twentieth of a point, so that is 36 pt — 12.7 mm. With a 20 mm margin, the paragraph's first word starts at exactly 32.7 mm, and its second line starts back at 20 mm.
The styling. "payable monthly" is drawn in Times italic, "first" carries an underline rule, and the comma straight after the italic run is upright — because the brace closed before it.
The tabs. "Term" and "Notice" are different widths, but the text after the tab on both lines starts at 32.7 mm. That's a tab stop doing its job. Swap tabs for a fixed run of spaces, as a lazy converter does, and the second column goes crooked by the difference between those two words.
Switch the typeface to sans and the page count stays at 1, but the body wraps in a different place — Helvetica is wider than Times, so the line breaks after "attracts" rather than after "at". The layout is measured, not guessed, which is why the page count you see is the page count you get.
How many pages you'll get
Page count is what people want to know before converting RTF to PDF, and it's decided by the settings — the .rtf has no say in the paper. Below is one 5,400-word, 40-paragraph agreement (28,560 characters) run through each, measured rather than estimated.
| Setting | Pages | Lines drawn | What it's good for |
|---|---|---|---|
| A4 portrait, 12 pt, 20 mm, 1.15 × | 7 | 320 | The defaults — letters, contracts, notes |
| Letter portrait | 7 | 280 | The same job on US paper |
| Legal portrait | 5 | 280 | Letter's width with three more inches of height |
| A4 landscape | 7 | 200 | Wide tabbed columns; no page saving |
| A4, base 10 pt | 5 | 240 | Long documents for printing |
| A4, base 14 pt | 9 | 360 | Reading on a phone, or older eyes |
| A4, double spaced | 12 | 320 | Manuscripts and anything to be marked up |
| A4, 10 mm margins | 6 | 280 | Getting under a page limit |
Two rows surprise people. Landscape gives far fewer lines — 200 instead of 320, each holding more words — and still costs seven pages, because a landscape A4 page is also much shorter. And dropping the base size from 12 pt to 10 pt saves two pages in seven, a bigger win than halving the margins.
The four numeric settings are bounded on purpose: base size 4–72 pt, margins 0–100 mm, line spacing 1–4 ×, paragraph spacing 0–72 pt. Type something outside a range and the box names the field and what it will accept.
Edge cases and gotchas
The apostrophe problem. A Word-written .rtf stores a curly apostrophe as the escape \'92 — a curly apostrophe in the Windows-1252 codepage, undefined in plain Latin-1. Decode it naively and every contraction in the document loses its apostrophe or grows a control character. Same story for \'93 and \'94, the curly double quotes. These decode properly here — a low bar that a surprising amount of software fails.
Your fonts aren't coming. Every PDF reader draws the same fourteen standard fonts with nothing embedded, and your document's font isn't one of them. So the tool reads its name out of the font table, tells you what it found, and draws in Times or Helvetica. Only the letterforms differ.
Soft hyphens are instructions, not characters. U+00AD marks a place where a word may break; draw it unconditionally and you get a stray hyphen mid-word. It's dropped, along with zero-width joiners and byte-order marks — invisible instructions rather than text.
Tables become tab-aligned text. Cell contents survive and land on tab stops; the rules and shading do not. A two-column schedule reads perfectly well. A dense financial table with merged cells will not, and that part is better exported to PDF from the application that made it.
The limits are real but large. 8 MB of RTF or 2,000 pages, whichever comes first. RTF markup inflates a file several times over the text inside it, so 8 MB is a genuinely long document — and if you pass either limit the page says which one rather than freezing the tab.
Related PDF tools
If the file you actually have is a .docx, it needs a different parser — the Word to PDF converter reads that format directly. Going the other way, back to something editable, is the PDF to Word converter, and pulling out just the words is Extract Text from PDF. For a plain text file with no formatting to keep — a log, a code listing, a column of numbers — the TXT to PDF converter uses a monospace grid instead, which holds columns in line the way a proportional face can't.
Once you have the PDF: PDF Merger joins it to other files, PDF Splitter pulls a page range back out, Sign PDF stamps a drawn signature onto a page, and Compress PDF shrinks it for email. All of them run in the browser, same as this one.
Frequently asked questions
Will the line breaks land in the same places as in my word processor?
Not always, and they can't. Line breaks depend on the exact widths of the exact glyphs drawn, and the PDF is drawn in Times or Helvetica rather than your document's font — a word that just fitted in Calibri may not fit in Helvetica. Everything structural is preserved: which paragraph the text belongs to, its alignment, indents, styling, and any page break the document asked for. If a paragraph has to end on a specific page, add an explicit page break before exporting — those are honoured exactly, and the page tells you how many it found.
Does it use the page size and margins set inside the .rtf file?
No — the paper is the one you pick. RTF can record a page setup, but it's usually whatever was on the machine the document was written on years ago, and it's the thing people most often want to change on the way to PDF. So page size, orientation, margins, line spacing and paragraph spacing come from the settings on this page. The file supplies everything else — the words, the styling, the alignment, the indents, the breaks.
What happens to headers, footers, footnotes and comments?
They're left out, on purpose. RTF stores each in its own named section, separate from the body, and a converter that doesn't recognise them dumps their contents into the middle of your document — a page number mid-sentence, a reviewer's comment pasted into a contract. Those sections are skipped, along with the document's metadata, so the body arrives clean. If you need page numbers on the finished file, add them afterwards with PDF Page Numbers.
Does it keep text colour or highlighting?
No. Every character is drawn in black, so colour is genuinely lost. If your document uses red text to mark changes or yellow highlighting to flag clauses, turn that into something structural before converting — bold, italic and underline all come through, and so does a marker word.
How do I get the page count down?
Try these in order, watching the count update as you type. Drop the base font size first — 12 pt to 10 pt took a 7-page document to 5. Then tighten line spacing toward 1. Then cut the gap between paragraphs, 6 pt by default and able to go to 0. Margins are the smallest lever: halving them saved one page and made the document look cramped. And if it's for a screen rather than a printer, try Legal paper — Letter's width with three more inches of height.
Can I convert several .rtf files at once?
One file at a time. Each document gets its own settings and its own PDF, named after the source — contract.rtf comes out as contract.pdf. If you want one combined document at the end, convert each file and join the results with PDF Merger. That's also the route for a document too long for one pass: split it, convert the parts, merge the PDFs.