CV / Resume Checker
Drop in your PDF and see it the way a recruiter’s software does — the sections it loses, the bullets that prove nothing, and the exact lines to change. Read in your browser, never uploaded.
How to use CV / Resume Checker
Drop in the PDF you actually send
Not the Word file, not an export made for this — the file a recruiter receives. Half of what is checked here is how that specific file parses, so a different export is a different result.
Pick the department you are applying to
This is the only required field, because it decides what counts as evidence. "Closed £1.4m against a £1m quota" is a strong bullet in sales and irrelevant on a clinical CV; a missing portfolio link is fatal in design and means nothing in accounting. Everything else — target role, career stage, the job advert — is optional and only sharpens the result.
Fix the critical findings before anything else
They are ordered by what they cost. A two-column layout or an unreadable text layer ends the application before a human sees it; a weak verb only costs you the read. Fixing the wording of a CV a parser cannot open changes nothing.
Open "what a parser sees" and read it
That pane is your file after extraction, which is what gets stored, searched and filtered. If a line is missing or scrambled there, it is missing for them — whatever the page looks like on screen.
Download the report and edit against it
Every finding quotes the line it came from and says what to write instead, so the report works as a checklist in whichever editor your CV actually lives in.
Why use this cv / resume checker
Runs on your device
Your input is handled entirely in this browser tab. Nothing is uploaded, so there is nothing for us to store, log or leak.
No sign-up, no quota
No account, no daily limit, no watermark and no email wall. Use it once or two hundred times a day.
Built to be linked
Options are stored in the URL, so a configured tool is a link you can send to a colleague or bookmark.
Technical notes
Most CV scorers show a percentage and sell the explanation. This one parses the PDF in your browser the way an applicant tracking system does — fonts, columns, text layer and all — then reports 43 named checks across seven weighted categories, judged against the department you are applying to, each one quoting the line in your CV that cost the points.
Measured on our own test set
- Checks run
- 43
- Departments
- 15
- Weighting
- parse 25 · impact 22 · structure 15 · keywords 15 · contact 10 · language 8 · length 5
- Parse gate
- caps the total
- Typical run
- under 1s
across parsing, contact, structure, evidence, keywords, language and length
each with its own expected skills, link, section and definition of a result
a CV a machine cannot read cannot average its way to a pass on wording
for a two-page CV, entirely on your device
Supported
- Department-aware scoring across 15 fields — the skills searched for, the link opened first, the section expected and what counts as a result all change with the field, because a sales CV and a clinical CV are not the same document scored twice
- PDF text extraction with no upload and no third-party library — objects, streams, fonts and content operators, so a parse failure is reported rather than quietly recovered
- Two-column detection, with a side-by-side view of how the gutter scrambles your CV when read straight across
- Hidden-text detection — white-on-white and invisible render modes, the keyword stuffing that gets applications discarded
- Unmappable glyph detection: fonts with no character map look perfect on screen and paste as boxes
- Section, role, date and bullet parsing, including gap detection and mixed date formats
- Bullet analysis: action verbs, quantified outcomes, duty-list phrasing, hedges, clichés, tense consistency and first person
- Keyword gap against a pasted job advert, matched on the advert's own wording
- Optional AI review whose rewrites are verified against your actual file before they are shown
Limits and trade-offs
- It cannot tell you whether you will get the interview. It reports whether the document is readable, findable and evidenced — the parts that are checkable.
- No tool knows a specific employer's scoring formula, because none of them publish one. Anything claiming a named system's score has invented it.
- Section headings are detected in English only. A CV in another language parses and scores on structure, but its headings will not be recognised.
- Scanned and image-only CVs are reported as unreadable rather than run through OCR — which is the finding, since that is exactly what happens to them downstream.
- Complex scripts that reorder glyphs — Bengali, Devanagari, Arabic, Thai — extract with imperfect word spacing, so the word count and language checks are approximate for CVs written in them.
- Career gaps, market-specific conventions and unconventional titles are reported as context, never scored as faults.
If something goes wrong
- This PDF is password-protected or restricted
- Remove the protection and export a plain PDF. An applicant tracking system cannot open a protected file either, so it would be rejected before anyone read it.
- No readable text — the file is an image
- Export straight to PDF from the original document instead of scanning or printing to image. If the original is gone, retype it: nothing recovers a CV that carries no text layer.
- Characters cannot be read as text
- Re-export with font embedding enabled, or switch to Arial, Calibri, Helvetica or Times. Test it by opening the PDF, selecting all and pasting into a plain text editor.
- Sections were not detected
- Use plain headings on their own line — Summary, Experience, Education, Skills — outside any table or text box. Creative section names cost you the field mapping.
Last verified August 2026 · benchmarks re-run each quarter.
Frequently asked questions
Is this a real ATS score?
It is a real measurement, and no tool can be more than that. No applicant tracking system publishes a scoring formula, and they differ — so a tool claiming "your Workday score" is inventing it. This measures the things every system depends on: whether the text extracts, whether the sections are findable, whether contact fields survive, and how the content compares to the advert.
Why does it only take PDF?
Because how your PDF parses is half of what is being checked, and a .docx cannot tell you that. The same CV exported two ways can differ by thirty points — one embeds its fonts and keeps a single column, the other loses a heading into a text box. The file you send is the file that has to be checked.
Is my CV uploaded anywhere?
No. The PDF is parsed inside this browser tab by code that ships with the page, so the file never leaves your device and closing the tab removes it. The optional AI review at the end is the only feature that sends anything, it asks before it runs, and it strips your name, email and phone first.
What score should I be aiming for?
Above 80, with nothing left in Critical. The number matters less than its composition: a 72 with clean parsing is a good CV that needs stronger bullets, and a 72 carrying a parse failure is a CV that will not arrive. Fix the named findings and the number follows.
Why did my beautifully designed CV score badly?
Almost always the layout. Two columns, tables and text boxes look considered and parse badly — a parser reads straight across the page, so a sidebar gets interleaved line by line into your job history. The "what a parser sees" pane shows exactly what that does to your file.
Does this work for non-technical CVs?
Yes — that is what the department selector is for. Pick your field and the checker changes what it expects: a nurse is checked for registrations and caseload, an account executive for quota attainment, a designer for a portfolio link, a researcher for publications. Only the parsing checks are the same for everyone, because a parser does not care what you do.
What does choosing a department actually change?
Four things: which skills a recruiter in that field searches for, which link they open first, which extra section is expected, and what a result is phrased as. The same front-end CV scores 83 as engineering and 80 as sales in our own testing — not because it got worse, but because it never states a quota, a pipeline figure or a win rate. Leave the field unselected and every department-specific check stays neutral rather than guessing.
Will the AI rewrite my CV for me?
It rewrites three to five bullets as examples, using only what your CV already claims, with square-bracket placeholders wherever a number is missing. It cannot invent your results — any rewrite whose "before" text is not genuinely in your file is discarded before you see it. The rest of the report is rules, and runs without it.
Does adding keywords from the advert actually help?
Covering requirements you genuinely meet, in the advert's own wording, helps — recruiters search shortlists with those exact terms. Stuffing them does not: it is obvious to a reader, and hidden white-on-white keywords are detected by every major system and treated as grounds to discard the application.