Guides
CV formatting software for recruiters: a 2026 buyer's guide
How to choose CV formatting software for a recruitment agency in 2026: accuracy, contact removal, client templates, DOCX vs PDF, pricing and a checklist.
CV formatting software takes a candidate's CV, usually a PDF or Word file, and turns it into your agency's branded submission for a client. The good tools do three things well. They keep every fact exactly as the candidate wrote it. They remove contact details from the whole file, hidden metadata included. And they give you an editable Word document plus a matching PDF in each client's format. This guide covers how to judge accuracy, contact removal, templates, file formats and pricing, and ends with a checklist for your trial. We build CVPitch, so we've said where that matters and kept the criteria general enough to use on any vendor.
What CV formatting software actually does
Most tools in this category follow the same pipeline:
- Intake. You upload a CV (PDF or DOCX, and sometimes scans or images).
- Extraction. The tool reads the text and pulls out employers, job titles, dates, education, certifications and skills. This is usually where AI comes in.
- Mapping. The extracted facts go into your template: agency header, section order, headings, fonts and page rules.
- Review. A recruiter checks the result and corrects it.
- Contact removal. Personal email, phone, address and links come off, depending on what the client should see.
- Export. You download a Word file, a PDF or both. Some tools also share a link or push the file back to your ATS.
Extraction and formatting are different jobs. If you're comparing formatters with the parser built into your ATS, read CV parser vs CV formatter first. A parser gives you searchable data. A formatter gives you a document you're willing to put your agency's name on.
The six criteria that matter
1. Factual accuracy you can check
Speed is easy to demo. Accuracy is harder, and it's what your client relationship depends on. A formatted CV that turns a part-finished degree into a completed one, or drops "contract" from a role, reflects on you, not the software.
In our October 2026 tests, three tools processed the same synthetic CV. It contained an MSc the candidate had withdrawn from, with "degree not awarded" written next to it. Two of the three showed the MSc with its dates but without the withdrawal note on the first pass, so it read like a completed degree. One kept the note. Both tools that dropped it put it back once we corrected the field. That's one test case, not an accuracy rate. It does show what to look for.
Ask every vendor:
- Can I see, for each extracted fact, the passage in the original CV it came from?
- What happens when the tool isn't sure about a date or a qualification? Does it flag it, or does it guess?
- Are things I add myself (notice period, rate, my own summary) labelled as recruiter additions, separate from the candidate's own content?
- Does the tool ever add skills, metrics or summaries the CV doesn't support?
There's more on how AI formatting changes facts, and how to catch it, in AI CV formatting accuracy.
2. Contact removal that covers the whole file
Removing the email from the visible header is the easy part. Contact details also hide in footers, text boxes, hyperlinks (the display text can say "LinkedIn" while the link holds a personal URL), document properties, the file name and the PDF's metadata.
Ask the vendor to show you the exported file, not the on-screen preview, with the contact details removed. Then check it yourself: open the DOCX, look at File > Info > Properties, and search the PDF's text. Our guide to removing contact details from a CV has the full manual check.
Removing contact details is also not the same as anonymising someone. A distinctive career history can still identify a person, and a good tool will say so.
3. Client templates without a designer on retainer
Agencies rarely have one format. Your banking client wants a one-page summary first. Your public-sector client wants a set section order and no photos. Ask:
- Can I keep a separate variant per client (logo, accent colour, section order, heading names, page limit)?
- Who builds the template, and is there a setup fee or a fee for each change?
- If a client changes its format, do older approved submissions still show the format they were sent in?
- Can the tool clone any Word template I upload, or does it work within a controlled set of layout options? Neither answer is wrong, but cloning arbitrary templates tends to cost more and break more often.
4. Editable DOCX and a matching PDF
| Format | Best for | Watch out for |
|---|---|---|
| DOCX (Word) | Last-minute tweaks, clients who want to annotate, ATS uploads | Fonts and page breaks can shift between Word versions and LibreOffice |
| Fixed layout, a polished look, email attachments | Hard to edit; metadata and hidden text can survive a careless export | |
| Both from one approved version | Client choice without the two files drifting apart | Edits you make later in Word aren't checked by the software |
If a tool exports PDF only, every correction means another round trip. If it exports Word only, you're relying on the client's Word setup to render it properly.
5. Review and approval controls
Look for a clear moment where a named recruiter approves a specific version. If someone edits a fact afterwards, or switches the template or privacy settings, the approval should no longer apply. Without that, the person who checked the CV may not be the person who sent it.
6. How candidate data is handled
Ask who can see uploaded CVs, how long they're kept, whether they're used to train models, which AI provider processes them and where, and how deletion works. Keep "we take security seriously" separate from things you can verify. Our security page sets out how CVPitch handles candidate data.
Turn your next CV into a client-ready submission
Upload a candidate CV, check every fact against the source, remove contact details and export your agency’s branded DOCX and PDF. Your first 10 CVs are free.
Pricing models compared
Pricing pages in this market use different units, so line them up before you compare.
| Model | How it works | Good for | Watch out for |
|---|---|---|---|
| Per CV or per export | You pay per formatted document | Low or irregular volume | Whether an edited re-export counts as a new CV |
| Monthly allowance | A set number of CVs per month | Steady volume | Unused allowance usually doesn't roll over |
| Per seat | Each recruiter is a licence | Small, stable teams | Costs grow with headcount, not usage |
| Quote-based | Sales-led, often with setup | Large agencies, complex integrations | Hard to compare; setup and template fees |
| Shared credits | Credits cover several tools | Teams using the wider suite | CV formatting may draw on the same pool as other features |
Pin down three things:
- What counts as one unit? An upload, a parsed CV, an export, or each re-export after an edit?
- Is setup extra? Template builds and later changes are sometimes charged separately.
- Are users capped? Seat pricing changes your sums when you hire.
Here's some illustrative arithmetic. If your desk sends 120 CVs a month, an allowance plan of 200 at $49 works out at about $0.41 for each CV you actually send. A per-CV price of $0.50 would cost $60. At 40 CVs a month the order flips. Run your real volumes through each model before you decide.
For reference, CVPitch's pricing is a free trial of 10 exports with no card, then Starter at $19 a month for 50 exports, Pro at $49 for 200 and Agency at $99 for 500. Every plan includes unlimited users. One export is one approved CV as a matching DOCX and PDF pair. Draft previews and re-downloads don't use an export.
How to run a fair trial
A vendor's demo CV is chosen to work. Your trial should include the CVs that cause trouble. Use synthetic CVs, or ones you have permission to use, and build a small test set:
- A CV with an unfinished or withdrawn qualification.
- A role with no end date, or only a year.
- A two-column PDF with a sidebar of skills.
- Contact details in the header or footer, plus a LinkedIn hyperlink.
- A negative statement, such as "no Python experience", which shouldn't turn into a Python skill.
For each file, time the whole task from upload to a file you'd send, including checking and fixing. Count the facts that went missing, changed or appeared from nowhere, separately from layout fixes. Open the DOCX in Word, not just the vendor's preview.
Buyer's checklist
- Each extracted fact links to its source passage in the original CV.
- Uncertain dates and qualifications are flagged, not guessed.
- Recruiter additions are labelled and kept apart from candidate facts.
- Contact removal covers headers, footers, hyperlinks, document properties and file names.
- There's a separate template variant for each client, and changing one doesn't rewrite past submissions.
- Exports include an editable DOCX and a PDF from the same approved version.
- Approval is tied to an exact version, and editing afterwards means approving again.
- The pricing unit is clear (upload, export or re-export), and you know whether setup is extra.
- Retention, deletion and access are documented, as is the AI provider used.
- There's a free trial long enough to test your awkward CVs, not just the demo.
Where CVPitch fits, and where it doesn't
CVPitch is built around the first criterion. Every extracted fact sits beside the passage it came from, uncertain items are flagged, and approval belongs to an exact revision, template and contact-removal policy. Contact removal covers the generated DOCX and PDF, hidden document metadata included.
It has limits you should know about. Files can be up to 10 MiB, with PDFs up to 20 pages. Scans and photos of CVs are read by OCR, but each page uses an AI credit and the transcription needs checking against the image. Nobody rebuilds your Word template for you: you pick one of five layouts in a client variant, covering branding, section order, headings, required sections, page limits and an optional coversheet, or add simple placeholder tags to your own Word template and upload it. Native connectors cover Greenhouse, Lever, Workable, Ashby, Teamtailor, Recruitee, SmartRecruiters, Recruit CRM, Crelate and Manatal; Bullhorn, JobAdder and Vincere are waiting on partner approval, so if you need one of those today, check that first. Everything else connects through the API, signed webhooks and automation tools such as Zapier, Make and n8n. For a wider look at the market, see our comparison of CV formatting software.
Frequently asked questions
What is CV formatting software?
It's software that turns a candidate's original CV into your agency's branded submission document. It extracts the facts, lays them out in your template, removes contact details where needed and exports the result, usually as Word and PDF.
Is AI CV formatting accurate enough to send without checking?
No tool should be trusted blind. In our tests, qualifications and date ranges were where meaning most often changed. Choose software that shows the source for each fact so that checking takes minutes rather than a side-by-side reread.
Should I send clients a Word file or a PDF?
Send whatever the client asks for. Many agencies send a PDF and keep the editable Word version for quick changes. The safest setup produces both from the same approved version, so they can't drift apart.
How much does CV formatting software cost?
Public prices range from per-CV pricing to monthly allowances, per-seat licences and quoted enterprise contracts. Compare the cost per CV you actually send, and include setup fees and whether re-exports are charged.
Bottom line
Pick the tool that makes your awkward CVs safe to send, not the one with the slickest demo. Test it with your own difficult files, check the exported documents yourself, and make sure the pricing unit matches how your desk actually works.
Turn your next CV into a client-ready submission
Upload a candidate CV, check every fact against the source, remove contact details and export your agency’s branded DOCX and PDF. Your first 10 CVs are free.