Dext vs Hubdoc for Bookkeepers Weighing Cost in 2026

Where the close actually stalls

Receipt capture is the step that turns a photo, email, or PDF of a bill into supplier, date, and amount fields a ledger can post. A bookkeeper searching this comparison is usually past the demo stage and is pricing one client file, or a stack of them, before the month closes. The category choice is depth of extraction versus a document inbox that already sits inside an eligible Xero subscription. Dext is the right shortlist when the file needs metered line items, bank-statement sheets, supplier statements, or a ledger list wider than the two products named on Hubdoc's pricing page. Hubdoc is the right shortlist when the client is already on an including Xero plan and header fields plus a stored image are enough to code the bill.

TL;DR: keep Hubdoc when the Xero plan already includes it and the bills are simple. Move the file to Dext when you will actually use the metered extraction, a non-Xero ledger, or practice controls that Hubdoc's setup guide does not describe. Neither product replaces the person who approves the account code.

The name on the older invoices matters. Receipt Bank was renamed Dext in 2021, and the flagship was then called Dext Prepare, at a time when the company reported 1 million users, according to Accounting Today (2021). That user count is a 2021 company report, not a 2026 customer census. IRIS Software Group, founded in 1978, completed its acquisition of Dext Software Ltd on 23 December 2024, according to IRIS (2024). The product a firm buys in 2026 is still sold under the Dext name on its own pricing pages. Hubdoc was not renamed. It remains a separate organization you connect to a Xero organization, and Xero's partner guide treats it as part of named Xero plans rather than as a discontinued app.

Key Takeaways

• Keep Hubdoc when an Early, Growing, or Established Xero plan already includes it and header fields (supplier, date, total, and invoice number) are enough to code the bill.

• Move the file to Dext when you will use metered line items, bank-statement sheets, supplier statements, a non-Xero ledger, or practice controls that Hubdoc's setup guide does not describe.

• Treat Dext practice Essentials and Advanced base rates as quote-based, with a 10-client minimum and 13 percent off for annual billing, and do not confuse that quote with the published business seat.

• Remember that Cashbook and Ledger bill Hubdoc separately, and that Early's 5-bill cap can stop a busy file even when Hubdoc describes unlimited usage.

• Expect Dext uploads past the 250-document business cap to wait for extraction until the next bill date or an upgrade.

• Leave auto-publish off in either product until a person has checked two clean bills from that supplier, because neither tool replaces account-code approval.

Who this is for

This comparison is for a bookkeeper inside an accounting firm who is about to put a client on a capture tool, or to move a client off one. You live in the publish step: the image has to become a bill, the supplier has to match the contact list, and someone has to notice a duplicate before it hits the bank feed. You are deciding for Xero files, QuickBooks Online files, or a mix, and you have to explain the software line on the client's bill.

Red flags: you need payroll, a sales-tax return, or a full ledger, which neither capture product is. You have not checked whether the client is on an including Xero plan or on Cashbook or Ledger before calling Hubdoc free. You need a finished per-client practice price and you have not built the quote, so any practice total in your head is a guess.

Choose Hubdoc for a client already on an including Xero plan whose bills can be coded from supplier, date, total, and invoice number, with the image attached for the reviewer. Choose Dext when that header set is not enough, when the practice publishes into more ledgers than Hubdoc's pricing page names, or when line-item splits, bank-statement sheets, and supplier-statement checks are real work this month. Disqualify Dext when the only pain is "we forget to forward receipts" and the Xero file is already inside Hubdoc with supplier rules. Disqualify Hubdoc when the client is not on an including plan and you also need the metered extraction Dext sells as named add-ons, because you would then pay Hubdoc's standalone rate and still lack that meter.

How we evaluated these capture tools

The weights below are a buying rubric for a firm bookkeeper, not a lab score and not a ranking taken from a review site. Cost of the capture seat is the heaviest weight because it is the line you either absorb or rebills. Extraction depth is next because line items and statements change how long coding takes. Ledger fit, plan caps, and practice scale split the rest, since a cheap inbox on the wrong Xero plan is not cheap, and a practice quote with a client minimum is a different product from a single-business subscription. Review and publish controls are the smallest weight because both products can auto-publish, and the failure mode is a human one either way. Weights sum to 100 percent. A file that misses the pass score on plan caps or ledger fit should not be moved on price alone.

Pass marks we would actually use: 15 of 25 on cost only if the marginal capture seat is zero or the metered extras are smaller than the coding time you can point at. 12 of 20 on extraction only if the file truly needs lines or statements. 9 of 15 on ledger fit if the target ledger is one both products name, or if Dext's wider list is required. 9 of 15 on plan caps if the host plan can hold the month's bills. 9 of 15 on practice scale if you have priced the real client count. 6 of 10 on publish controls if a named person still sees exceptions before approval.

What each product still is

Dext today is a bookkeeping capture and publish platform. On the practice feature list it includes multi-channel receipt and invoice capture, WhatsApp, intelligent extraction, supplier rules, supplier-statement reconciliation, approvals, duplicate detection, line-item extraction, bank match, missing-paperwork requests, auto-publish, employee expense management, and mileage tracking. The same page prices practice plans per client, with a 10-client minimum and 13 percent off for annual billing, and it does not print a public base rate for the Essentials plan or the Advanced plan. Published add-ons on that page include custom guidance at $7 per client, Data Health and Insights at $7.5 per client per month, and Commerce Lite at $7.5 per client per month, and the page states integrations with 36+ accounting solutions, according to Dext (2026). Treat the Essentials and Advanced base rates as quote-based until that builder returns a number for your client list.

Hubdoc today is a document inbox that extracts key data and publishes a transaction into Xero. The partner guide names collection by desktop upload, a unique intake email, the mobile app, ScanSnap, and multi-page or split PDFs. It says Hubdoc extracts contact, date, total amount, invoice number, and more, imports contacts and the chart of accounts, and can apply supplier rules with auto sync so a known supplier posts without a fresh coding session. After publish, the guide points you at the Xero bank feed for the match. Xero includes Hubdoc on Early, Growing, and Established, bills it separately on Cashbook and Ledger, gives a Xero partner a practice Hubdoc account, and suggests bringing clients on in groups of 5, according to Xero (2026). The same guide says there is no cost to add staff to the practice organization, then you grant those people client by client. Each client organization is created from that client's Xero organization and connects to it.

Dext review score: 4.3 from 175 reviews and Hubdoc at 4.2 from 92 reviews, with value-for-money scores of 3.9 and 4.1, on a comparison updated 1 October 2026, according to Capterra (2026). The gap is small. Read it as a satisfaction snapshot, not as proof of extraction accuracy. Capterra also shows a higher value score for Hubdoc, which matches the bundle story, and you should still price the host Xero plan before you treat that score as a saving.

Side-by-side on the jobs you code

The matrix uses only behavior described on the pages opened for this guide. Where a page is silent, the cell says so. Silence is not a promise that the feature is absent.

A firm with only Xero clients can live in the right-hand column for a long time. A firm that also publishes to QuickBooks Online can stay with Hubdoc's standalone product, because that logo is on Hubdoc's pricing page, but the include-with-Xero benefit does not follow the file to QuickBooks. Dext's 36+ claim is the wider door, and it is a count of connections, not a promise that every field maps cleanly. Test the account code, the tax rate, and the attachment on one supplier before you turn auto-publish on for the whole client.

Pricing checked against the public pages

Pricing checked October 10, 2026. Annual business rate: $25.21 per month ($302.50 billed annually) for 250 documents and 5 users, with 10 bank-statement sheets, 5 line-item documents, and 5 supplier-statement extractions included. Line-item and supplier-statement pay-as-you-go start at $0.50 per document, bank-statement pay-as-you-go starts at $0.32 per document, those bundles start at $20.50 and $13.00 per month, AI Assist is $10.5 per month until 31 January 2027, Commerce Lite starts at $7.50 per month, and Vault is free to 100MB then from $4 per month. The same page says over 700,000 businesses and 12,000 accounting and bookkeeping firms use Dext, and that past the document cap you may still upload but extraction waits until the next bill date or an upgrade, according to Dext (2026). The rendered page emphasized the annual figure and said annual billing is up to 20 percent cheaper. This guide does not print a monthly business rate that was not shown as a dollar figure on that render. The 14-day trial is stated on the page.

Standalone rate: $12 per month after a 30-day trial with no card required, with unlimited usage, multiple collaborators, and sync to Xero and QuickBooks Online named on the page, according to Hubdoc (2026). Unlimited usage is Hubdoc's phrase for its own product. It is not a promise that the connected ledger will accept an unlimited number of bills.

The Early plan on Xero is $27 per month after $5.40 for the first 3 months, the Growing plan on Xero is $59 per month after $11.80, and the Established plan on Xero is $97 per month after $19.40. The page offers 80 percent off for 3 months to eligible new US customers on a first organization, and it caps Early at 20 invoices and 5 bills, according to Xero (2026). Growing and Established list automated bill entry without a numeric bill cap on that page. Growing shows a 60-day cash-flow forecast and Established shows 180 days. If a client is on Early, confirm whether bills published from Hubdoc count toward the 5-bill cap before you move a busy file there. A $0 Hubdoc seat on a plan that stops at 5 bills is not a high-volume design.

The next table is the same published figures in numeric cells so a partner can sort a file without rereading the paragraphs. The plan cell says what the count means. Paired dollars are both published. They are not a discount we calculated.

Practice add-ons stay in prose so they are not mistaken for the business seat: $7 per client for custom guidance, and $7.5 per client per month for Data Health and for Commerce Lite. A 10-client minimum means a three-client firm should compare the business seat per entity with a quote before assuming the practice SKU is cheaper. No per-client base dollar is printed here because none was printed on the page.

One client month, counted out

Take a single client that forwards 180 supplier documents in a month, 12 of them itemized bills you actually split. The business plan cited above includes 250 documents and 5 line-item documents, so 7 documents sit outside that line-item allowance. Seven times the cited $0.50 pay-as-you-go rate is $3.50, and the cited $25.21 monthly rate plus $3.50 is $28.71 before tax. The cited $12 standalone rate on Hubdoc is lower, and on the Growing plan the capture seat is already inside the cited $59 subscription, so the software delta versus Hubdoc-on-Growing is the full $28.71 rather than a gap between two capture apps. The illustrative tie-out is to read annual.billsCount on the annual object in Dext's Data Health activity-stats response and ask whether that yearly ledger count is even in the neighborhood of 180 times 12, which is 2,160, before anyone treats the inbox and the books as the same queue. Documents forwarded are not the same object as posted bills, so a mismatch is a review item, not an error by itself. This arithmetic is an illustration from published rates, not a timed study and not a client result.

Dext for files that outgrow header capture

Best fit is a business already near the cited allowance, or a firm whose clients are not all on including Xero plans. Five users on the business seat covers a small client team plus the bookkeeper. The included 5 line-item documents and 10 bank sheets are a sample, not a month of itemized payables. If every supplier invoice must be split, price the $0.50 documents or the bundle that starts at $20.50 before you tell the client the seat is $25.21. Multi-entity businesses past five entities are pointed at a bespoke conversation on the pricing page, so do not assume the slider price survives that shape.

Limits are the cap and the quote. Past 250 documents, extraction stops even though upload does not, so a seasonal client can look "in the tool" while the inbox is full of images with no fields. Practice base rates are quote-based, and the 10-client minimum is a real gate for a small book. Data Health at $7.5 per client per month is optional. It is also the product surface that exposes client activity fields such as annual.billsCount. Do not buy it only to decorate a dashboard. Buy it if a reviewer will use those counts.

Implementation, from the public plan rather than from a project we ran: open the 14-day trial, connect the ledger you already use, and send one supplier through mobile or email before you write rules for the rest. Set supplier rules only after the account code and tax rate match the ledger on a real bill. Keep auto-publish off until a second bill from that supplier lands clean. If the firm is pricing the practice SKU, build the quote at the true client count, then add only the Data Health or Commerce Lite seats you will open. The human review point is the item that is missing a category, a supplier match, or a line split. That item should wait. Cancel paths differ by term: the pricing FAQ says a monthly plan can be cancelled before renewal, and an annual plan needs 30 days' notice before renewal.

Hubdoc for Xero files that stay simple

Best fit is a client on Growing or Established whose suppliers repeat, who will email or photograph bills, and whose coding is stable enough for a supplier rule. The bookkeeper's win is the image on the transaction and a contact that already exists in Xero, not a fully split bill. QuickBooks Online remains on the pricing page for firms that are not on Xero, at the cited $12 rate, with unlimited usage in Hubdoc's own wording. Partners are told to stand up their own practice organization first, learn the product, then add clients in the groups of 5 the guide suggests, which is an adoption pace, not a technical limit.

Limits sit mostly in the host plan and in the field list. The guide's named extract is contact, date, total amount, invoice number, and more. Anything past that list, including a reliable line split, was not given a price or an allowance on the pages opened, so do not sell it as included. Early's 5-bill cap can end the design even when Hubdoc says unlimited. Cashbook and Ledger do not include the seat. Disconnecting Hubdoc from an including plan is a billing event you should read in Xero's help before you promise a client they can leave and keep extracting. One Hubdoc organization connects to one Xero organization in the setup the guide describes, so a client with two Xero files needs two Hubdoc organizations, not a toggle.

Implementation follows the guide's clicks, not a custom project. From the client's Xero organization, open Hubdoc, accept the terms, and create the organization. Invite the client only when you want them inside. Add firm staff at the practice level, then grant the clients they may see. Agree a weekly capture habit: photos, forwarded email, and a scan if they use ScanSnap. Write supplier rules for the vendors that repeat, and leave new suppliers in a state a person codes. The review point is a new supplier, a total that does not match the image, or a bill that would push an Early file through its cap. Publish after that check, then reconcile on the Xero bank feed as the guide describes. The guide also says a Xero partner's practice Hubdoc account is included, which is the right place to practice rules before a client sees them.

A review queue above the publish step

Capture tools stop at a published bill. The mess that remains is the exception: a missing due date, a supplier that matched the wrong contact, a duplicate image, or a month whose document count and ledger count tell different stories. A proposed workflow from US Tech Automations can sit after publish. The trigger is a new bill arriving in the ledger, or a nightly export the firm already downloads. The prerequisite is access the firm controls: ledger credentials or that export, a supplier map the firm maintains, and, only if you use Dext Data Health, an API token created in Practice settings for the Data Health product. The action is to compare supplier, total, and date with open bills, and to read annual.billsCount only as the reasonableness check in the illustration above. The output is an exception list. Nothing is marked approved by the workflow.

US Tech Automations would hold an item when the due date is blank or the supplier is new, and would assign it to a named bookkeeper instead of inventing a date or a contact. The same document identifier is stored so a replay does not create a second bill. That idempotency check is a configuration, and it depends on the export actually containing a stable id. If the file has no stable id, the design stops and a person matches the image. The review point stays on tax, account code, and any line split. This is a proposed design, not a live deployment and not a measured saving.

Zapier, Make, and n8n can run this shape. Configured with care, they keep run histories, retries, error branches, and an audit of what fired. The buyer owns observability, idempotency, escalation, access control, and the maintenance when a ledger field changes. A proposed US Tech Automations design can pre-build the exception queue, the duplicate check, and the escalation to a named reviewer, with the same prerequisites: credentials or an export you already have, a supplier map, and a person on the ugly items. Those automation tools are a fair choice when the firm already staffs them and the volume is a handful of rules. They are a poor place to hide an unowned close.

When NOT to use US Tech Automations

Skip an extra queue when the client is on Growing or Established, Hubdoc supplier rules already post clean header bills, and the only leftover work is the bank match your staff already do in Xero. Skip it when one business is inside the Dext allowance, supplier rules are stable, and a person already clears the inbox before publish. Skip it when the firm wants a one-month cleanup in a spreadsheet and will not keep credentials, reviewers, and failure alerts current. In those three cases the capture tool plus the ledger is the smaller system, and adding a workflow would only add a login.

Mistakes that make a cheap inbox expensive

The usual miss is treating "included" as "every Xero plan." Cashbook and Ledger are called out as billed separately. The next miss is parking a heavy bill volume on Early because Hubdoc's own page says unlimited usage, while Early lists 5 bills. The third is comparing the business seat with a practice quote as if they were one SKU. They are not. The business seat has published dollars and a document cap. The practice base is quote-based and has a 10-client minimum. A fourth miss is turning on auto-publish, in either product, before two clean bills from that supplier have been checked by a person. A fifth is copying a review site's starting price into a proposal. Use the vendor page cited above, and use the same dollar every time you repeat it.

Before you move a client, write down the host plan, the month's document count, how many of those documents need lines, and which ledger receives the publish. If you cannot fill those four blanks, you are not ready to pick. The same habit of checking the source before you recommend a tool shows up when clients ask about other software. A dental client comparing in-house plan platforms can start from this membership-platform comparison. A SaaS client comparing writing tools can start from this SaaS writing-tool comparison. An agency client who needs a month-end warning on spend can read this underpacing recipe. Those are adjacent buying problems, not capture features.

Terms worth agreeing before you switch

• Receipt capture: turning a source document into fields a ledger can post, plus keeping the image.

• Header extract: the top of the bill, such as supplier, date, total, and invoice number, as Hubdoc's guide names them.

• Line-item extraction: a priced Dext allowance that splits a document into lines, with 5 documents included on the business plan.

• Supplier rule: a saved coding instruction for a repeating vendor, available in both products' public descriptions.

• Auto-publish: sending the coded document through without a fresh review. Both products describe a version of this. Leave it off until rules are proven.

• Document allowance: Dext's monthly cap, 250 documents on the business plan cited here. Uploads past the cap are not the same as extracted bills.

• Including plan: Early, Growing, or Established, the Xero plans whose guide says Hubdoc is included. Cashbook and Ledger are outside that list.

• Quote-based: a practice base rate that was not printed as dollars on the page opened for this guide.

Questions bookkeepers ask before they move a file

Is Dext still Receipt Bank?

Yes. Accounting Today reported the 2021 rename from Receipt Bank to Dext, with the flagship then called Dext Prepare. The pricing pages reviewed for this article sell the product as Dext, and IRIS completed the acquisition on 23 December 2024. Your old login name may linger in a client's head. The subscription you are pricing is the current Dext seat, business or practice, not a legacy Receipt Bank SKU with a different allowance.

Is Hubdoc included with every Xero subscription?

No. The partner guide includes it on Early, Growing, and Established, and says Cashbook and Ledger are billed separately at Hubdoc's retail price. A Xero partner's own practice organization is described as included. Confirm the client's plan name in Xero before you strike the capture line off the engagement letter. If the plan is Early, also read the 5-bill cap, because inclusion does not raise that cap.

What does a single business pay for Dext?

The US business page shows $25.21 per month on annual billing, $302.50 for the year, for 250 documents and 5 users, with the small included allowances for lines, bank sheets, and supplier statements cited above. Past the document cap, extraction pauses. Extras are the pay-as-you-go rates or the bundles, not a silent overage on the base seat. A 14-day trial is offered. Practice pricing is a different page and is quote-based at the base.

Can both products send bills to QuickBooks Online?

Yes, on the pages reviewed. Hubdoc's pricing page names QuickBooks Online next to Xero. Dext's practice page says 36+ accounting solutions, which is broader than two logos and still needs a field test on the actual file. The include-with-Xero benefit does not apply to a QuickBooks-only client. That client either pays Hubdoc's $12 rate or pays Dext. Publishing still depends on a person checking supplier, tax, and account code unless a proven rule is already on.

What happens when a Dext file passes 250 documents?

The pricing FAQ says you can keep uploading, and data is not extracted until the next bill date or you upgrade. For a client with a seasonal spike, that is the operational risk, not the $25.21 sticker. Either raise the plan before the spike or accept a queue of images that a person will key. Do not discover the pause on the last day of the month. The 180-document illustration earlier stays inside the cap. A 300-document month does not.

When should the firm stay on Hubdoc?

Stay when the client is on Growing or Established, the guide's header fields are enough, supplier rules cover the repeating vendors, and nobody is asking for metered line items or bank-statement sheets. Also stay, at the $12 rate, when a QuickBooks Online file is light and you do not need Dext's wider practice controls. Move when Early's 5-bill cap blocks the month, when the plan is Cashbook or Ledger and you would pay retail anyway, or when the coding requires the line-item meter often enough that the $0.50 documents are the real product you are buying.

Does a higher review score mean cleaner bills?

No. The Capterra snapshot cited above is 4.3 from 175 reviews for Dext and 4.2 from 92 reviews for Hubdoc, with Hubdoc ahead on value for money at 4.1 versus 3.9. That does not measure your suppliers, your tax codes, or your Early cap. Use the scores as a check that neither product is an outlier in public reviews. Use a live supplier bill as the acceptance test. If the first ten bills need heavy correction, the score will not rescue the close.

What to do with the file you have

Price the host plan first, then the capture seat. If the client is on Growing at the cited $59 rate or Established at the cited $97 rate and the bills are header-level, Hubdoc is the included path and Dext's $25.21 seat needs a reason you can show, such as line items, another ledger, or statement extraction. If the client is on Early, model 5 bills before you praise unlimited capture. If the firm wants Dext across many clients, get the quote. Do not invent a per-client dollar the page did not show. Keep a person on new suppliers and broken totals whichever product you pick.

If the remaining pain is the exception list after publish, see how US Tech Automations configures this as a queue with credentials, a stable document id, and a named reviewer. If that pain is not there, leave the stack as the capture tool plus the ledger and spend the hour on supplier rules instead.