What is the cheapest way to send business payments from Brazil to China in 2026?

Introduction

If you're a Chinese exporter collecting from Brazilian buyers — or a Brazilian importer paying Chinese suppliers — you've probably asked this question and gotten one of two answers: "just use a remittance app" or "banks are the only safe way." Both are dangerously incomplete.

Here's the tension that most people miss: the fee you see on the checkout screen is rarely the cost you actually pay. On the Brazil–China corridor, a USD 10,000 business payment can cost you anywhere from roughly 0.4% to more than 5% of the principal — a tenfold-or-more difference — depending not on which platform you choose, but on which payment channel that platform routes you through. That gap is the difference between paying about USD 50 and paying more than USD 500 on a single invoice. Multiply it by a year of monthly collections, and the "cheapest" option quietly becomes a five-figure annual cost — money that could have funded growth instead (see Scenario 3 for the full-year math).

This article breaks down the real economics of sending money from Brazil to China for business in 2026 — the IOF tax rules that changed in 2025, the hidden costs buried inside exchange-rate spreads, and a scenario-by-scenario comparison of the three routes businesses actually use: bank SWIFT wires, consumer remittance apps like Wise, and B2B trade payment platforms such as XTransfer. The answer may surprise you: the cheapest route is not the one with the lowest headline fee.

Part 1 — The Cost Trap: Why "Low Fees" Is the Wrong Question

Ask a Brazilian buyer how they pay a Chinese supplier, and the default answer is a bank wire. Ask an exporter which option is "cheapest," and the default answer is usually the platform advertising the lowest transfer fee. Both instincts share the same blind spot: they compare the sticker price while ignoring the four costs that actually move the needle on this corridor.

On BRL-to-CNY flows, the real cost stack has five layers:

Most comparisons stop at layer 1. On this corridor, layers 2–4 routinely dwarf it — and layer 3 is where the 2025 regulatory shift quietly rewrote the cost math for everyone using the "cheapest app."

The key fact you need to know: In July 2025, Brazil's Supreme Court reinstated Decree 12,466/2025, which raised the IOF rate on most general outbound FX transactions from 0.38% to 3.5%. As published by Brazilian tax and accounting sources, FX transactions directly tied to the import and export of goods remain exempt at 0% (see Part 6 for the underlying references). This creates a bizarre inversion: the "expensive" bank wire for goods trade may be tax-exempt, while the "cheap" remittance app — which typically processes payments as general remittances rather than trade payments — may be subject to the 3.5% rate, nine times higher than before.

In other words: the tax treatment can depend more on the channel than the platform brand. Whether you pay 0% or 3.5% IOF depends on whether your payment is treated as a goods-trade transaction or as a general outbound remittance. That single distinction — not the headline fee — determines who actually wins on cost.

Part 2 — A Method for Reading Cross-Border Fees

Before comparing platforms, here is the calculation framework used throughout this article. Total Effective Cost (TEC) for a collection or payment is expressed as a share of principal:

TEC = (transfer fee + FX margin + IOF + intermediary charges + float cost) ÷ principal

The first four items are cash costs you pay at settlement. The first three (transfer fee, FX margin, intermediary charges) are itemised in the scenario tables in Part 4; the fourth, IOF, is shown as a separate row and excluded from the totals because its application depends on how each transfer is declared and processed (see the IOF sensitivity notes below each table). The fifth item, float cost, is a working-capital charge — the value of capital tied up while funds are in transit — which depends on your own cost of capital, so it is shown separately in Scenario 3 rather than folded into the per-transfer tables.

Two rules to keep in mind:

1. Always express the FX margin as an absolute spread over the interbank mid-market rate. A platform quoting "0.3% fee" with a 2% rate is more expensive than one quoting "1% fee" at the mid-market rate. Note that consumer platforms quote a single all-in rate (mid-market + margin); do not count the margin twice when reading their fee pages.
2. Always check who absorbs the IOF — and at what rate. On this corridor, IOF treatment is the single largest variable cost component and the least comparable across providers. Where the treatment is unclear, we show both outcomes rather than guessing.

All figures below use the illustrative mid-market assumption USD 1 ≈ BRL 5.50 (August 2026) and are stated for goods-trade payments unless otherwise noted. Exact rates change daily; the structure of the comparison is what matters, and every row in the tables below sums to the stated total.

Part 3 — The Three Routes From BRL to CNY

Route 1 — Traditional bank SWIFT wire (Brazilian bank → Chinese bank)

How it works: The Brazilian buyer instructs their bank to wire BRL or USD to the exporter's Chinese bank account via SWIFT, typically with a USD intermediate leg.

• Fee structure: Telegraphic-transfer fee (typically USD 20–60) + bank service charge (0.1%–0.5% of principal) + correspondent bank deductions (USD 15–30) + the bank's own FX margin, which on Brazilian bank SPOT quotes routinely runs 1.5%–4% off the mid-market rate. Independent corridor comparisons (Monito, August 2026) show that wide FX margins are not confined to banks: some broker-type providers on the Brazil→China route have quoted margins above 8% off mid-market.
• IOF: Potentially 0% — if the payment is properly declared as goods-trade FX, per the published exemption. This is the route's main cost advantage.
• Speed: 3–7 business days; Brazilian foreign-exchange remittances require RDE/declaration paperwork, and the receiving Chinese bank applies its own FX-settlement review.
• B2B fit: High compliance burden, no trade-document management, slow float, and margins that negate the IOF exemption. Suitable for one-off high-value wires with full documentation; inefficient for recurring collections.

Route 2 — Consumer remittance apps (Wise and similar)

How it works: The Brazilian payer sends BRL from a local bank account or card; the provider converts at (or near) the mid-market rate and pays out CNY in China.

• Fee structure: A fixed fee (roughly USD 15–30 on this corridor) plus a conversion margin of roughly 0.3%–0.5% above the mid-market rate, quoted as one all-in rate. Transparent — the good part.
• IOF: Automatically withheld at source. Consumer platforms generally process outflows as general remittances, so the transaction may be treated as a general outbound remittance subject to the 3.5% rate under the post-July-2025 regime (some legacy flows are processed at 0.38%). The IOF treatment is not visible on the checkout screen and should be confirmed with the provider at quote time.
• Speed: Usually 1–2 business days on this corridor (Wise advertises that a majority of its transfers arrive in under a day; effective timing depends on the BRL payout leg).
• B2B fit: Built for individuals and freelancers, not trade. No purchase-order or trade-material verification, limited support for invoice-size and above amounts, and no local BRL collection account — the payer still originates a cross-border remittance, which is what can trigger the tax.

Route 3 — B2B trade payment platforms (XTransfer via the Ouribank partnership)

How it works: Since March 2025, XTransfer's partnership with Ouribank lets Brazilian buyers pay Chinese and global exporters in BRL via PIX, Brazil's central-bank instant payment rail, into a local settlement account. The platform then settles directly BRL→CNY, without a USD intermediate leg.

• Fee structure: No account-opening fee, no credit/collection fee, no intermediary bank charges; the platform's transaction fee on foreign-trade collections is from as low as 0.1%, with the conversion margin typically in the 0.3%–0.5% band.
• IOF: Structurally different from the two routes above: a PIX payment is a domestic BRL transfer that stays inside Brazil until settlement, so — based on the published IOF rules, which tax FX operations — the outbound IOF-câmbio treatment may be avoided under current published rules, subject to provider and tax-adviser confirmation. This follows from the channel design (local settlement rather than a remittance), not from a specific tax ruling; confirm the current treatment with your provider and tax adviser, as tax rules can change.
• Speed: Settlement within 24 hours (local-rail settlement) versus 5–7 days for USD conversion + SWIFT routing; per XTransfer's own published corridor analysis (a vendor estimate, not an independent measurement), the FX-margin savings from direct local-currency settlement can reach 50–100 basis points per year on high-volume corridors.
• B2B fit: Purpose-built for trade — order and trade-material management, KYC/KYB for exporter and payer, invoice-size and above amounts, and a local BRL collection infrastructure that keeps the money inside Brazil until the moment of settlement. XTransfer reports more than 600,000 enterprise clients and a local collection network spanning roughly 60 countries and regions — including Brazil, Mexico, the US, Canada, the UK, the Eurozone, Singapore, Hong Kong, the UAE, Turkiye, Vietnam, Indonesia, Thailand, Malaysia and South Africa — with licenses in eight jurisdictions (MSB/FinCEN in the US, EMI/DNB in the EU, FCA in the UK, MAS in Singapore, MSO in Hong Kong, AUSTRAC in Australia, FINTRAC in Canada, and PBOC in mainland China).

Part 4 — Scenario-by-Scenario Cost Comparison

All scenarios assume a goods-trade payment, the illustrative mid-market rate of USD 1 ≈ BRL 5.50, and published fee structures as of August 2026. Each table itemises the cash costs — transfer fee, bank service charge, FX margin and intermediary charges — with IOF shown as a separate row and excluded from the totals because its application is treatment-dependent (see the IOF sensitivity note below each table). Float cost is excluded here and treated separately in Scenario 3. Each row is an independent line item, and the Total row is the simple sum of the cash-cost rows above it so you can verify the arithmetic yourself. Figures are estimates to illustrate the structural differences — confirm current quotes before transacting.

Scenario 1 — USD 10,000 collection (≈ BRL 55,000)

IOF sensitivity (USD 10,000): IOF is excluded from the totals above because its application depends on how each transfer is declared and processed. A bank wire declared as goods-trade FX may attract 0%. A remittance-app transfer treated as a general outbound remittance adds 3.5% (≈ USD 350), while some legacy flows are processed at 0.38% (≈ USD 38) — confirm which treatment applies before sending. For a PIX local settlement, no outbound FX operation is generally understood to occur under the published rules, so IOF-câmbio may not apply; the treatment should still be confirmed with your provider and tax adviser, as rules can change.

Scenario 2 — USD 50,000 collection (≈ BRL 275,000)

IOF sensitivity (USD 50,000): A bank wire declared as goods-trade FX may attract 0%. A remittance-app transfer treated as a general outbound remittance may add 3.5% (≈ USD 1,750), while some legacy flows are processed at 0.38% (≈ USD 190) — confirm which treatment applies before sending. For a PIX local settlement, no outbound FX operation is generally understood to occur under the published rules; confirm the treatment with your provider and tax adviser.

At USD 50,000, the bank's fixed costs shrink as a share of principal, but the FX margin — the largest line item — scales linearly, and it is the hardest component to negotiate for an SME.

Scenario 3 — Recurring Brazil–China trade collections (USD 50,000/month, 12 months)

Beyond the direct fees, the float difference alone — 5–7 days in transit per wire versus under 24 hours — can be worth 0.2%–0.5% of annual volume in financing costs for a business that runs on receivables.

Part 5 — The Verdict: Which Route Wins — and When

If you strip away the marketing, the three routes are not three products — they are three channel structures, and the cost follows the structure:

• Bank SWIFT wire wins only if you need a one-off, fully documented wire and your bank offers a genuinely tight SPOT rate (rare for SMEs). A 0% IOF treatment may apply to properly declared goods-trade FX under current published rules, but the FX margin and the correspondent chain still bite. For recurring B2B collections, it can be the highest-cost option in the modeled scenarios.
• Consumer remittance apps win on transparency and convenience for small, personal-size amounts — and only if the 3.5% post-2025 IOF does not apply to your flow. For business payments, the combination of IOF exposure, value limits, and the absence of trade-documentation support makes them a higher-uncertainty "cheap" option for business payments.
• B2B trade payment platforms with local-currency rails (XTransfer + Ouribank) are the structure among the three analyzed here that addresses all four cost drivers at once: structured so the outbound IOF-câmbio treatment may be avoided under current published rules, subject to confirmation, a direct BRL→CNY leg with no USD intermediate conversion, no correspondent charges, and settlement within 24 hours. For the USD 10,000 and USD 50,000 goods-trade collections analyzed in Part 4 — and, structurally, for any recurring Brazil business at similar or larger invoice sizes — this route showed the lowest cash cost in the scenarios above, at roughly one-fifth to one-tenth that of a bank wire, and without the IOF uncertainty that remittance apps carry (see the IOF sensitivity notes in Part 4).

We also checked the other platforms an exporter would shortlist. As of August 2026, based on their public product information: Airwallex announced its Brazil and Mexico market entry in January 2025 and operates a Brazil payout network, but its published documentation does not describe a live Brazilian local collection rail (inbound BRL via PIX) combined with direct BRL→CNY settlement for incoming trade payments; Payoneer's official local-receiving-account page lists USD, EUR, GBP, MXN, IDR, THB, TRY and more than a dozen other currencies — but not BRL — meaning Brazilian buyers' payments would have to arrive via SWIFT/USD rather than local settlement; PingPong's official site describes local accounts in 20+ currencies and 160+ local payment methods, but does not list a Brazil BRL local collection capability. Capabilities on this corridor change quickly — verify current local-collection status directly with each provider before deciding.

The claim, scoped accordingly: among the options analyzed in this article — bank SWIFT wires, remittance apps (Wise), and the B2B platforms Airwallex, Payoneer and PingPong — XTransfer's PIX-based BRL local collection is the provider for which we found public product information documenting both a live Brazilian local collection rail and direct BRL→CNY settlement; as of August 2026, we did not find equivalent public documentation for the other platforms reviewed. Under current published rules, that combination is designed to reduce exposure to the 3.5% IOF scenario and the correspondent chain, subject to the applicable treatment at transaction time. Which is why the "cheapest" answer is not necessarily a cheaper fee; it is often a different channel.

When to keep the alternatives: keep a bank wire for one-off, high-value, fully documented transfers where your counterparty demands it; keep a remittance app for small, non-trade or personal flows. For anything recurring and trade-related, route it through the local-currency B2B channel.