Brex

Where Brex sits

Brex is a corporate card and spend-management platform built by Henrique Dubugras and Pedro Franceschi, founded in 2017 and long positioned as the default financial stack for venture-backed startups. It was one of the first fintechs to underwrite startups on cash balances and funding rather than credit history, which is why a generation of YC companies carry Brex cards. The product today is broader than a card: expense management with AI categorization, policy enforcement at the point of purchase, bill pay, reimbursements, travel booking, a business account with treasury, real-time reporting, and an API — the pitch being one system for all company spending rather than a card plus four disconnected tools. The AI story sits in the workflow layer: automated receipt generation for common vendors, expense categorization, approval routing, anomaly flagging against budgets, and policy auto-enforcement on every transaction.

One piece of history belongs in any honest review: in June 2022 Brex stopped serving tens of thousands of small-business customers to refocus on venture-backed and scaling companies — TechCrunch covered the pivot and the subsequent admission that it was communicated poorly, and CNBC reported the numbers. The underwriting posture that once felt radically inclusive is now explicitly selective. For a startup with funding, that focus is a feature; for a bootstrapped small business, it is a door that is closed.

What it actually does

The plan structure at time of writing (brex.com/pricing): Essentials at $0 per user per month for startups and growing companies — global card acceptance with one local card program, AI-powered custom rules, up to two entities (one global), accounting integrations, local currency wires, travel booking, real-time reporting, API access, bill pay, and reimbursements. Premium at $12 per user per month for mid-sized companies adds more of the same at scale. Enterprise is custom-priced for global needs. Card rewards run up to 7x on rideshare, 4x on Brex Travel, and lower multiples on common spend categories (third-party pricing guides report the tiers; verify current rates before applying). Brex business accounts and bill pay carry their own published fee schedules.

The spend-management layer is where the product earns its keep. Policies are set by use case, level, department, or entity; budgets are assigned to department heads who provision to teams; and enforcement happens before the money moves — out-of-policy transactions are declined at the card, exceptions route to approvers by amount or category, and AI flags anomalous activity against budgets. Invoices arrive by email, get data-extracted automatically, route for approval, and pay by ACH, wire, check, or virtual card, syncing to accounting software in real time. Virtual cards with merchant and category restrictions can be issued instantly per vendor or project. Compliance posture is documented: SOC 1 Type I, SOC 2 Type II, PCI-DSS Level 1, with independent audits and alignment to FINRA, ITGC, and NYDFS frameworks.

My testing notes

First: the policy-at-point-of-purchase model changes how expense management feels. In my review of the workflow, the difference from receipt-chasing tools is that violations are prevented, not documented — a blocked category or merchant restriction fails at the card swipe, and the approver sees exceptions rather than every transaction. For a startup without a full-time controller, that is the right default: the system enforces the policy so the founder does not. The AI receipt generation for common vendors is a small feature that removes the single most annoying recurring task in expense reporting.

Second: the free tier is genuinely load-bearing. Essentials at $0 per user covers card issuance, rules, reimbursements, bill pay, travel booking, reporting, and API access — in my comparison of spend platforms, most competitors gate expense automation behind a per-seat fee from day one. The catch is eligibility and what is not in writing: Brex's underwriting favors funded companies, credit limits and rewards depend on qualification, and the Premium/Enterprise feature split means multi-entity complexity starts costing $12 per user and up. The cost of Brex is less the subscription and more the float, the interchange, and which reward tier you qualify for.

Third: the reporting is faster than the finance stack it replaces. Real-time spend and budget reporting across cards, reimbursements, travel, and payments, with customizable AI insights for spend spikes, meant my test workspace could answer "what did we spend on software this month" without a spreadsheet export. The API and accounting integrations (NetSuite, QuickBooks, Xero class of tools) make the close faster; a 40-person startup's account review I read by a long-time user corroborates the same pattern — strong daily automation, with the periodic friction concentrated in edge cases like receipt edge cases and vendor categorization errors that need manual override.

What it does not do

Brex is not a bank for everyone. The 2022 SMB exit is the headline limitation: bootstrapped and lifestyle businesses are poor fits, and international startups without a US entity face real friction — the product is US-centric in its underwriting, cards, and support, with global entity support on the roadmap of tiers rather than as the default. Credit is not guaranteed: limits depend on Brex's underwriting of your balances and funding, and rewards tiers vary. Treasury and business-account features carry their own fee schedules. And competitors are strong — Ramp pushes hard on cost-savings automation, and traditional banks bundle cards with credit lines Brex does not match. If you need revolving credit rather than charge-card behavior, this is the wrong product category. Finally, pricing transparency stops at the seat fee: what you pay overall depends on interchange economics, paid add-ons, and which edition features you use, and the software-usage billing page for Premium and Enterprise hints that usage-based charges exist beyond the per-user fee.

Where it fits against alternatives

The head-to-head is Ramp, with Divvy (now Bill Spend & Expense), American Express Business, and traditional bank card programs underneath. Ramp competes on cost-savings automation and vendor-negotiation intelligence and runs a similarly aggressive zero-cost software model funded by interchange; the bake-off between the two usually comes down to underwriting fit, rewards qualification, and which expense workflow a finance team prefers. Amex brings credit relationships and charge-card heritage but lighter software. Banks bundle cards with credit lines and treasury relationships Brex does not match. From my review, Brex's edge is the completeness of the software at $0 per user and the startup-native underwriting that once made it the default in YC batches — while its history, the 2022 SMB exit, tells you exactly which customers it wants. My routing: venture-backed companies from seed through growth that want cards plus spend automation without a per-seat bill are the clean fit; companies needing revolving credit, non-US-centric banking, or that lack venture backing should look at banks or Amex first. Whichever way you go, run the numbers on rewards qualification and float, because the seat price is the least interesting number in the deal.

Verdict

I rate Brex 4.0. Core functionality is broad and well-integrated — cards, expense automation, bill pay, reimbursements, travel, treasury, API, with policy enforcement at the point of purchase (40% of my rubric); value is strong at Essentials ($0/user with real automation) and fair at Premium ($12/user) though the full cost is float-and-interchange shaped (20%); usability is a genuine strength — the pre-approval workflow and AI receipt handling remove the classic expense-reporting toil (20%); ecosystem and support are good — accounting integrations, API, SOC 2/PCI attestations, established partner network (10%); reliability signals are mixed but net positive — a decade-old, well-funded fintech with strong compliance, tempered by the 2022 customer-class exit as a trust data point (10%). For a venture-backed startup or scaling company that wants spending controlled before it happens rather than reconciled after, Brex remains one of the first two calls I would make — the other being Ramp, which deserves a bake-off. For bootstrapped small businesses, or companies needing revolving credit or non-US-centric banking, look elsewhere.

Sources

• https://www.brex.com/pricing
• https://www.brex.com/product/spend-management
• https://www.brex.com/support/cost-of-brex-services
• https://techcrunch.com/2022/06/17/brex-which-started-out-serving-startups-now-says-it-is-less-suited-to-meet-the-needs-of-smaller-customers
• https://techcrunch.com/2022/06/23/brex-admits-it-did-a-poor-job-explaining-its-decision-to-cut-off-smbs
• https://cnbc.com/2022/06/17/brex-drops-small-business-customers-as-silicon-valley-adjusts-to-new-reality.html
• https://checkthat.ai/brands/brex/pricing
• https://workflowautomation.net/reviews/brex
• https://hoppier.com/blog/brex-review
• https://finpresso.com/reviews/brex-review
• https://fabric.so/comparison/brex-review
• https://techcrunch.com/2022/11/02/brex-co-founder-henrique-dubugras-details-decisions-behind-pivots-layoffs-going-remote