MineralTree vs Bill.com for AP Automation in 2026

Key Takeaways

• MineralTree is the business-to-business payables business of Global Payments, not a Billtrust receivables product. The 2021 sale price on the advisor note was $500 million, and a June 4, 2025 product release still uses that ownership.

• The name on the old search, Bill.com, is the company that rebranded to BILL in 2022. You are comparing a live payables platform with a live payables platform, not a discontinued brand.

• BILL publishes three seat prices and a custom enterprise tier, plus a per-payment fee card. MineralTree's TotalAP page describes setup, an annual platform fee, and monthly usage fees, and does not post the dollars.

• Automatic sync to Oracle NetSuite and Sage Intacct is on BILL's custom Enterprise column. The TotalAP page lists both of those ledgers on its own connector list, and a separate Sage Intacct screen pays vendors without leaving Intacct.

• BILL's own pages do not agree on network size. The about page shows 8.3 million members in one block and over 4.7 million in the FAQ, and the pricing page says +4M in plan text and 8M+ in a table.

• A useful short list is ledger first, then whether you can accept a quote, then whether custom approvals and dual control force the enterprise tier.

TL;DR: pick BILL when QuickBooks Online or Xero sync and a public seat-plus-rail price are enough, and pick MineralTree when NetSuite or Sage Intacct is the system of record and you will negotiate a volume fee.

Accounts payable automation is the software that captures a supplier invoice, routes approval, and sends payment without someone retyping the bill into the accounting system.

Bill.com is now BILL, and MineralTree is not Billtrust

Controllers type the old product name because that is what colleagues still say. BILL's own timeline marks a 2022 rebrand from Bill.com to BILL, and the about-page FAQ describes the change as a move from a bill-pay utility to a platform that also covers spend and receivables. The company still trades under the ticker people already know. You are not shopping a sunset SKU. You are shopping the current accounts payable plans on the pricing page, which still sell invoice intake, approvals, and payments.

MineralTree did not move to Billtrust. Billtrust is a separate accounts-receivable business. No page opened for this comparison shows Billtrust as the buyer or the operator of MineralTree. William Blair advised the sale of MineralTree to Global Payments for $500 million and, in that October 22, 2021 note, said the company then served over 2,200 businesses and processed roughly 17 million invoices a year, according to William Blair (2021). Treat the 2,200 figure and the 17 million invoices as 2021 deal context, not as a 2026 customer count. On June 4, 2025, Global Payments described MineralTree as its business-to-business division and introduced Vendor Payments powered by MineralTree inside Sage Intacct, with ACH, virtual card, and check from the customer's bank, according to MineralTree (2025). That embedded screen is a payments layer inside Intacct. It is not the same offer as full invoice capture.

BILL's scale block lists Annual payment volume: $345B+ according to BILL (2026), while the FAQ on that same page says the network connects over 4.7 million members as of 2026 and the scale block says 8.3 million network members. The footnote on the 98 of the top 100 US accounting firms line is dated March 2026. Those are three different counts on one about page, before the pricing page adds +4M in the plan cards and 8M+ in the comparison table. Do not put a single network size in your business case. Ask BILL which count is current for vendors you can pay without collecting bank details, and ask what happens when the supplier is outside that network.

The category around both products is crowded, and a wider map of accounts payable automation software is the right place to see tools this pair does not cover. If the real contest is card-first payables rather than a classic invoice workflow, the BILL and Ramp accounts payable comparison and the BILL, Ramp, and Brex automation comparison sit beside this one. A nonprofit ledger has its own constraints, covered in the nonprofit accounting AP automation guide.

How we evaluated MineralTree against BILL

The reader is a controller or AP lead close to a signature. The job is invoice-to-pay for a company that already has an accounting system, a bank, and a set of approvers. Marketing adjectives do not decide that purchase. Sync, a price you can put in a spreadsheet, capture quality, approval depth, how money leaves the bank, and the audit trail do.

We weighted the six criteria below and refused a seventh that often sneaks into these write-ups: a star rating. Review-site scores for these two products were not on pages that loaded cleanly in this session, so there is no score row. A vendor claim of time saved is kept only with its year. Ardent Partners' 2026 benchmarks, as published by Basware from research with nearly 200 finance and AP leaders, put Best-in-class invoice cost: $2.67 according to Basware (2026), against $12.50 for all others. That gap is an industry reference for why capture and exceptions dominate the weight. It is not a measured result for either vendor, and it is not a promise that buying one of these tools produces the $2.67 figure.

Sync is 25% because a mid-market team will not replace NetSuite, Sage Intacct, or QuickBooks to suit a payables add-on. Capture and duplicates are 20% because that is the daily keystrokes and the duplicate-payment risk. Posted price is 15% because a quote you receive after a demo cannot be compared with a fee card on the same day. Approval depth is 15% because custom policies and a second control on payment are what push buyers from a listed tier into a custom contract. Rails are 15% because per-item ACH and check fees, or a usage fee inside a volume quote, move the year more than a small seat difference. The audit trail is 10% because both vendors talk about electronic history, and the practical test is whether a person still releases exceptions.

A missing answer is a disqualifier, not a tie. If the ledger is NetSuite or Sage Intacct and the buyer will not take a custom enterprise contract, BILL's published Corporate price does not cover that sync. If the buyer will not proceed without a public dollar price, MineralTree drops out until a quote arrives. If the pain is customer invoicing and cash application, neither product in this pair is the Billtrust-style receivables buy.

The seven decisions that change the contract

The feature matrix uses only what the pricing page, the TotalAP page, the June 2025 release, and BILL's bill API reference actually show. "Not published" means the page opened for that vendor did not state the item. It is not a claim that the feature is absent in a private contract.

Under that matrix, the operational hole is the exception, not the happy path. A proposed configuration from US Tech Automations watches a new BILL bill or a scheduled export from the payables file. The trigger is a repeated invoice number for the same vendor, or an amount over a limit the controller writes down. The action holds payment and assigns a review task to a named AP lead. The output is a review record an auditor can open later. Prerequisites are a BILL developer key, a session from that vendor's login, and a person who is allowed to release the hold. This is a design you would configure. It is not a live customer, and it has no measured time saving.

BILL's create-bill reference shows the invoice number nested on the bill, and it shows that a duplicate can be rejected only when the caller asks for rejection. Leaving the default in place will not save you. MineralTree's TotalAP page lists duplicate detection among payment controls, alongside role-based access, dual approvals, invoice flags, two-factor authentication, and positive pay. It does not publish the field name a programmer would call. For that product, the practical prerequisite is the file or the connector you already use, not an assumed public object.

Sync is the other fork. On the TotalAP page, API connections are named for Sage Intacct, QuickBooks, NetSuite, and Microsoft Dynamics GP. The same page says Xero is supported, and Sage 50 plus Acumatica go through a CSV file sync called File Connect. Multi-entity is described as native for NetSuite or Sage Intacct with top-level sync, and as multiple companies for QuickBooks, Dynamics GP, Xero, and File Connect. BILL's pricing table is stricter and more public. Essentials is manual CSV for QuickBooks Online, QuickBooks Pro, QuickBooks Premier, and Xero. The Team plan and the Corporate plan add automatic two-way sync for those products. QuickBooks Enterprise, Oracle NetSuite, Sage Intacct, Microsoft Dynamics, Acumatica, and Rillet appear on the Enterprise column, with a note that matching features vary by accounting software. API access is marked included on every BILL plan in the extras table, including Essentials. Do not let a salesperson tell you the API itself is the reason to buy Enterprise. Dual control, premium phone support, and those mid-market ledgers are the Enterprise reasons.

Approvals follow the same split. The TotalAP page says invoices route by your rules, with email reminders. It does not publish a plan grid. BILL includes standard approval policies on the listed plans, and custom approval policies start on the Corporate plan. Approver-only discounts are mentioned for Corporate and Enterprise, without a percent on the page, so the discount is quote-based even when the seat is not. Procurement with two-way matching is included on Corporate. Essentials and Team show that procurement block as an add-on, and the add-on price is not on the page. Three-way matching is described on the Enterprise card, again with the variance note.

Payment source matters for fraud and for the rebate conversation. The June 2025 MineralTree release says Vendor Payments powered by MineralTree debit the customer's existing bank and support virtual-card rebates. The TotalAP page says direct-debit payments are how bank matching stays simple. BILL's fee card prices payments funded from a bank, a BILL balance, or a BILL cash account, and it prices a separate card-funded path. Those are different cash paths. A controller who must keep every supplier payment on the operating account should get the funding source in the order form, not in a slide.

Pricing checked against the pages vendors publish

Pricing checked October 9, 2026.

The Corporate plan on BILL is Corporate seat price: $89 per user/month according to BILL (2026). The Essentials plan on BILL is $49 per user per month, the Team plan on BILL is $65 per user per month, and the Enterprise plan is Custom Pricing, so that seat is quote-based. The same page prices ACH at $0.59, a mailed check at $1.99, a virtual card as free, an international USD wire at $19.99, Pay Faster ACH at $11.99, and the Spend and Expense plan at $0 per user per month. A 50% time-saved line on that page is footnoted to a 2021 survey of over 2,000 customers. Use the 2021 label if you repeat it. It is not a 2026 benchmark. Plan cards on that page say +4M vendors. The comparison table on that page says 8M+ vendors. Keep both phrases tied to the block you saw, and do not average them.

The TotalAP page states Capture accuracy claim: up to 99% according to MineralTree (2026), from optical character recognition plus human validation. That page says the price includes one-time setup fees, an annual platform fee, and monthly usage fees for invoice capture, ACH payments, and check payments, with extra integration fees possible. Average implementation is 4 weeks from kickoff to first payment, and some clients are onboarded in 2 to 3 weeks. Security documentation is described as typically available within 7 to 10 business days of a request. The same page names SOC 1 Type 2, SOC 2 Type 2, PCI DSS Level 1, SOX, HIPAA, and GDPR among the audits and rules it says it maintains. Capture at "up to 99%" is a vendor claim with a human still in the loop. It is not a guarantee on your invoice layouts.

Read the fee language twice before you model MineralTree as "unlimited users, no rail fees." The specific product page opened here itemizes monthly usage fees for capture, ACH, and checks. Older all-inclusive wording still appears elsewhere on the site. Ask the seller which sentence the order form follows, and ask for the usage rate per invoice, per ACH, and per check in writing. Until that quote exists, the honest cell is quote-based.

BILL's payor fee card is public, which is why a seat-only comparison understates the year. These are the listed payor charges, not a MineralTree schedule.

International FX wire is labeled free on that page, with an exchange rate. Virtual card is labeled free. Neither free label removes the seat price. Corporate and Enterprise can discount approver-only users, but the discounted dollar is not printed, so do not invent it. Spend and Expense at $0 per user per month is software only. Card and payment fees still apply where the page says they do. Credit lines from $1,000 to $5 million are advertised and explicitly not guaranteed.

A total-cost model for a year is seats times months, plus each rail times its fee, plus 1099 forms if you file them here, plus any overnight checks. MineralTree's model is the quote: setup, plus the annual platform fee, plus monthly usage, plus integration fees if the ledger or the bank needs them. You cannot subtract one from the other until the quote arrives. Virtual-card rebates, if you earn them, belong in that model only after the vendor writes the rate. Neither page opened here states a universal rebate percent.

Which team each product actually fits

Choose the Team plan on BILL when the ledger is QuickBooks Online, QuickBooks Pro, QuickBooks Premier, or Xero, and you want automatic two-way sync, the invoice coding agent, and custom user roles without custom approval policies. Choose the Corporate plan on BILL when you also need custom approval policies, single sign-on, multi-entity, and procurement with two-way matching, and your ledger is still in that QuickBooks or Xero set. Choose the Enterprise plan on BILL when the ledger is NetSuite, Sage Intacct, Microsoft Dynamics, QuickBooks Enterprise, Acumatica, or Rillet, or when you need dual control and priority support. The Essentials plan on BILL fits a team that will live with CSV and standard roles. It is a weak fit for a multi-entity close.

Choose MineralTree when the operating picture matches the TotalAP page: invoice capture with a person checking the read, approvals by rules you already use, and payments debited from your bank, including a virtual-card path you want priced as a rebate rather than as a per-item ACH fee. The fit is stronger when Sage Intacct or NetSuite is the ledger, because that page lists those connectors and describes native multi-entity sync, and because the June 2025 release puts a thinner payments experience inside Sage Intacct for teams that do not want a second login. Choose the embedded Intacct payments screen, not full TotalAP, when invoices are already coded in Intacct and the only gap is paying them. Choose full TotalAP when capture and approval are still outside the ledger.

The same Basware summary of Ardent Partners puts Best-in-class approval time: 4.1 days according to Basware (2026), against 15.4 days for all others, with best-in-class teams moving 55.2% of invoices through with no manual touch versus 31.4% for all others, and with 58% of AP organizations using or piloting AI. Use those four figures as a yardstick for your own cycle time. Do not paste them into a vendor scorecard. Neither product page opened here publishes your future cycle time.

Limitations are specific. MineralTree will not give you a public seat price or a public per-ACH dollar on the pages opened, the implementation is measured in weeks, and a programmer cannot point to a public bill field the way BILL's API does. BILL will charge every full user on Essentials and Team, will not put NetSuite or Sage Intacct sync on the $89 tier, and publishes conflicting network counts. The invoice coding agent is absent on Essentials. Custom approval policies are absent below Corporate. Default duplicate handling on the API allows a repeated invoice number unless you turn rejection on. The 50% claim is a 2021 survey.

US Tech Automations can also be configured for the close, which neither fee card solves by itself. The trigger is a nightly payment file from MineralTree, from BILL, or from the accounting system. The action matches cleared ACH and check lines to open bills and flags any line with no bank reference. The output is an exception list for the person who owns the reconciliation, not an automatic journal entry. Prerequisites are the export the ledger already produces, a field map the controller signs, and a rule that unmatched lines stay open. No deployment count and no hours saved are claimed for that design.

A priced month, using only listed fees

Picture a controller running an illustrative month of 6 full-feature seats on the Corporate plan, 200 ACH payments, and 40 mailed checks, using only the fees published on the pricing page: 6 times $89 equals $534 in seats, 200 times $0.59 equals $118 in ACH fees, and 40 times $1.99 equals $79.60 in check fees, for a listed-fee month of $731.60 and a twelve-month extension of $8,779.20 if that mix never changes. On a create-bill call, the documented invoice object carries invoice.invoiceNumber, and a repeated value for the same vendor is rejected only if the request says to reject it, which still leaves a person to release any held bill before money moves. MineralTree's side of the same month stays quote-based, because the TotalAP page does not list a seat price or a per-payment dollar fee to multiply. If even 40 of those 240 payments had to go overnight, the BILL card adds 40 times $24.99, which is $999.60, and the month is no longer the $731.60 figure. That is why the rail mix belongs in the same spreadsheet as the seats.

The industry cost per invoice is a different math problem. At the $12.50 all-others figure from the Basware summary of Ardent Partners, 240 invoices would be $3,000 of process cost before any software fee. At the $2.67 best-in-class figure, the same 240 invoices would be $640.80. The difference, $2,359.20 in that illustration, is larger than the $731.60 software-and-rail month above. It is still not evidence that either vendor delivers the lower number. It is the reason a controller should time 25 of their own invoices before believing a vendor slide. The $2.67 and $12.50 figures stay exactly those values wherever they appear in this piece. They are 2026 benchmark citations, not this company's books.

Who this is for

This comparison is for a controller or AP lead who will sign an invoice-to-pay contract and already knows the accounting system, the bank, and the approvers. It is for a team that can separate a public fee card from a volume quote, and that will ask which legal entity issues virtual cards after any parent-company restructuring. It is also for the person who must explain the choice to an auditor without inventing a payback period.

Red flags: you need accounts receivable and cash application rather than supplier payments, which is a different vendor category than this pair. Your ledger is NetSuite or Sage Intacct and you refuse any contract that is not on a public price list, because BILL puts that sync on a custom tier and MineralTree does not publish the dollars. You need a programmer's field guide for MineralTree bill objects before signing, because the pages opened here do not publish one.

A short checklist before the call: name the ledger and the entities, count full users versus approvers, pull last month's ACH, check, card, and wire counts, list the invoices that were duplicates or exceptions, and decide who may release a held payment. If any of those five is unknown, you are not ready to compare the two quotes.

Building the same control in a general connector

The real alternative for many teams is not a third payables suite. It is a route built in Zapier, Make, or n8n, or a small internal job that reads a CSV. Those tools can keep run histories, retries, error branches, and audit evidence when someone configures them. They do not arrive with your chart of accounts, your approval matrix, or your bank's positive-pay file. The buyer has to design observability, idempotency so the same invoice is not paid twice, escalation when a step fails, access control on the connection, and the maintenance when a vendor changes a column.

A proposed US Tech Automations design would configure the duplicate hold and the nightly match as named steps with a required human release, still depending on the BILL API or on a file export from MineralTree or the ledger, and still stopping when the reviewer has not acted. A general connector can do those same steps. The difference you are buying, if you buy a configured design at all, is that the review stop and the field map are specified up front, with the prerequisite credentials written down. You still own the bank and the accounting system. You should still reject any proposal that skips the human release on an exception.

When NOT to use US Tech Automations, keep the work inside the payables product you already run. If every approver lives in BILL or in MineralTree and nobody needs a second queue, another system adds a login and a failure point. If a Sage Intacct team will pay only through Vendor Payments powered by MineralTree and will not export exceptions, the embedded screen is the smaller tool. If an administrator wants to change routes every week in a general connector and does not want a written review rule, a heavier design is the wrong weight.

Terms that show up in the order form

Invoice-to-pay. Capture, approval, and payment in one path, as opposed to a bank portal that pays a bill you already keyed.

ACH. An electronic bank payment. BILL lists $0.59 on the standard rail and $11.99 on the faster rail. MineralTree's per-ACH dollar is inside the quote.

Virtual card. A single-use card number for one supplier payment. BILL lists it as free to send. MineralTree ties it to rebates and to a direct debit from your bank. Get the rebate rate in writing either way.

Two-way sync. Invoices and payment status move both directions so the ledger and the payables tool do not diverge. Confirm which objects sync. A logo on a slide is not a sync scope.

Three-way match. The invoice is checked against the purchase order and the receipt. On BILL's page this is an Enterprise description, and it varies by accounting software. Do not assume the Corporate plan's two-way match includes the receipt.

Dual control. A second person must confirm a sensitive action. BILL lists dual control on Enterprise only. MineralTree lists dual approvals as a payment control without a public tier. Ask which actions are covered.

Quote-based. The vendor will not show the dollar until you ask. MineralTree's platform, setup, and usage fees are quote-based. BILL's Enterprise seat and its approver-only discount are quote-based. The ACH and check fees on BILL's card are not.

Positive pay. The bank receives a list of checks you intended to issue, and it flags anything else. MineralTree names it among controls. Confirm your bank participates before you drop paper controls.

Mistakes that waste the evaluation

• Treating Billtrust as MineralTree's owner. The public record opened here points to Global Payments, including the 2021 sale and the June 4, 2025 release.

• Assuming the Corporate plan price includes NetSuite or Sage Intacct. Those ledgers are on the Enterprise column.

• Modeling MineralTree as zero transaction fees because an older sentence sounds all-inclusive. The TotalAP page names monthly usage fees for capture, ACH, and checks.

• Comparing seat prices and ignoring 200 ACH payments. At $0.59 each, that rail is $118, before any check fees.

• Quoting the 50% time-saved line without the 2021 survey footnote, or quoting $2.67 per invoice as if it were your future cost.

• Collapsing 8.3 million, 4.7 million, +4M, and 8M+ into one network statistic. They are four different phrases on BILL's own pages.

• Skipping the duplicate control. A field that exists, including invoice.invoiceNumber, does nothing until the create call rejects repeats and a person reviews the hold.

• Buying a second workflow tool when the embedded Sage Intacct payments screen already covers the only gap you have.

Questions controllers ask before the short list

Is MineralTree owned by Billtrust?

No. Public pages identify MineralTree with Global Payments, including a 2021 acquisition note and a 2025 product release, and Billtrust is a different receivables company.

Is Bill.com gone?

No. The company rebranded to BILL in 2022 and still sells accounts payable, receivables, and spend on the current site. The search name is the old brand, not a discontinued product.

Does MineralTree publish a price?

No. The product page describes setup, an annual platform fee, and monthly usage fees, so the dollar amount is quote-based until the seller writes it down.

Which BILL plan includes NetSuite or Sage Intacct sync?

The Enterprise plan. The published Team and Corporate plans list automatic two-way sync for QuickBooks Online, QuickBooks Pro, QuickBooks Premier, and Xero, not for those two mid-market ledgers.

Can either product file a 1099 by itself for a posted fee?

BILL lists 1099 e-file to the IRS at $2.99 per form on the pricing page. MineralTree's pages opened here do not publish a 1099 filing fee, so confirm that step before you drop your current process.

What should we ignore in a vendor time-saved claim?

Ignore any percent that has no year and no sample. The 50% line on BILL's pricing page is tied to a 2021 customer survey, and the $2.67 invoice benchmark is an industry figure, not a result from your file.

The choice in one sitting is ledger, then price visibility, then who releases an exception. If you want the review stop and the nightly match specified as configurable steps, with the API or file prerequisite and a human release written down, see how US Tech Automations configures the review stop. Bring the seller a month of real payment counts either way. A quote without those counts is not a comparison.