Agentcard
Agentcard issues spend-capped virtual Visa cards so AI agents can complete real online purchases with human approval.
If your agent needs to complete a checkout today, Agentcard solves the hard part: a card scoped to one store and one amount, a human approval gate, and 1% revenue share that flows back to you. The free Vault tier (5,000 users) plus a $0.15-per-request Purchase Agent that's free in beta makes it cheap to prove one flow before you commit to $5,000/mo Issuing. The caveat is maturity — SOC 2 and PCI DSS sit behind the Enterprise tier, and the merchant network is a fixed list of named stores.
Verified 4d ago · liveness 82/100 · cite: rightaichoice.com/tools/agentcard
- Consumer apps whose agents purchase on a customer's behalf and need a stored card
- B2B teams automating company spend with a Slack or internal agent
- Agent-browser and agent-hosting platforms adding payment capability to their sessions
- Crypto products that turn user balances into online spend via USDC on Base
- Teams requiring on-premise or air-gapped deployment
- Merchants outside the covered store list unless you supply your own browser automation
- Buyers who need a completed SOC 2 and PCI DSS review in hand before signing
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Agentcard if you need a completed SOC 2 or PCI DSS report today, cannot commit to the $5,000/mo Issuing plan, or your agent's target merchants fall outside the 18+ covered stores.
Issuing runs $5,000/mo as a flat fee, and 1% of every settled transaction is taken as revenue share — that 1% scales with volume in a way the flat fee hides.
Free Vault under 5,000 users is genuinely cheap to evaluate, and Purchase Agent at $0.15/request (free in beta) lets you ship one merchant flow without a card-issuing contract. Issuing at $5,000/mo plus 1% revenue share is priced for companies with real agent checkout volume — a serious commitment next to a hobby-scale Stripe Issuing setup, and below enterprise card-issuing contracts that usually require far larger minimums.
In short
Agentcard — Agentcard issues spend-capped virtual Visa cards so AI agents can complete real online purchases with human approval. Best for Consumer apps whose agents purchase on a customer's behalf and need a stored card, B2B teams automating company spend with a Slack or internal agent, Agent-browser and agent-hosting platforms adding payment capability to their sessions. Free to start; paid plans from $0.15.
What's new in Agentcard
Checked 5 days agoAcross the latest 8 updates: 2 feature updates and 6 changelog entries.
Agentcard shopping agent refreshes expired approval links
The shopping agent now re-checks out when an approval link expires, returns the order already placed, and reports why a merchant rejected a delivery address.
Agentcard fixes partial card-network releases keeping residual holds
Partial releases now credit only the released amount and wait for final settlement, rather than treating the purchase as fully reversed.
Agentcard wallet feed adds authorization adjustment rows
Wallet activity now shows Spend adjusted rows and partial credits when a hold drops, instead of omitting the released amount.
Agentcard: 12-digit cards now support auto-approval and shared budgets
Auto-approval now accepts 12-digit cards issued by some networks. Payments at merchants whose own checkout encrypts the card from their servers remain refused.
Agentcard auto-approval switch consolidated into shared pool
The auto-approval panel no longer duplicates the switch on Your cards; toggling the shared pool restarts a period with nothing spent.
Agentcard refuses webhook addresses pointing at loopback in IPv6 form
Company webhook addresses resolving to loopback or internal addresses in IPv6 form are now rejected, matching existing plain-form behavior.
Agentcard blocks live payments to processor test platforms
Live payments now refuse a processor's test platform, and checkouts whose host doesn't match the environment are turned away before approval.
Agentcard 3D Secure window reports verification outcome
The 3D Secure verification window now states whether the bank check completed, failed, or was canceled, and adds a close button.
What people actually say about Agentcard — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
29 mentions across 3 sources (Hacker News, YouTube, Product Hunt) · researched Aug 2, 2026.
Average across the 3 sources that answered — each source counts once, not each post.
- +Single-use cards auto-cancel after one payment — perfect for agent safety
- +Human approval via notification adds a necessary control layer
- +Supports major agents like Claude, ChatGPT, Cursor via MCP and CLI
- +Direct purchases from DoorDash and Good Eggs via agent prompt
- +Dashboard tracks spending across agents — good for auditability
- −Security bug with test API keys allowed unauthorized purchases
- −No clear behavior when payment fails — does card retry or expire?
- −Limited community feedback and real-world testing evidence
- −One-time cards may not work for merchants with pre-auth holds
- −No mention of support for multi-currency or international transactions
- • Transaction fees per card may apply
- • Partner purchase fees (DoorDash, Good Eggs) could be added
- • Overage charges if you exceed plan limits
Viability Score
How well maintained and how widely used is Agentcard? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Single-use virtual Visa cards destroyed after one payment
- Multi-use Agentcards with spending caps for recurring subscriptions
- Cards locked to one merchant and one amount at issuance
- Human tap-to-approve via passkey or master password before payment
- Voice approval: customer speaks a passphrase with no screen involved
- Vault stores customers' existing cards encrypted on the device
- Rewards routing automatically picks the customer's highest-earning card
- Purchase Agent covers 18+ merchants including Amazon, Walmart, Target and DoorDash
- Purchase Agent confirms the exact cart and returns changed prices before paying
- MCP buy tool so Claude, Cursor or a custom agent can call the Purchase Agent
- Browser integrations for Kernel, Browserbase, Browser Use and Anchor
- Works with your own Playwright/Chromium via CDP attach
- iMessage and messaging agent payments via Linq and Blooio
- Tokenback pays rewards as Anthropic and OpenAI tokens
- USDC on Base for holding funds
About Agentcard
Agentcard is payment infrastructure for companies whose agents need to buy things online. Rather than hand an autonomous agent your credit card, you issue it a virtual Visa locked to one merchant and one amount, then a human taps approve (passkey or master password) before any money moves. The stack splits into four products: Vault stores customers' or employees' existing cards — encrypted on the device — and lets an agent buy with them; Issuing mints one-time and multi-use cards from a prefunded balance; Purchase Agent runs Agentcard's own shopping agent across 18+ named merchants; and Voice lets a customer approve a purchase by speaking a passphrase back, with no screen involved. The distribution strategy is where this differs from generic virtual-card APIs. Browser integrations for Kernel, Browserbase, Browser Use, Anchor and your own Playwright/Chromium pull Agentcard into whatever session the agent already drives, so the card is attached at checkout rather than bolted on afterward. Messaging integrations with Linq and Blooio do the same for iMessage-native agents — Linq sends the approval as a native bubble, then hands the agent a card. The Purchase Agent works with any agent over one API call or the buy tool over MCP, which means Claude, Cursor or a custom loop can buy without you running a browser stack at all. Pricing reflects that split. Vault is free up to 5,000 users and takes 1% revenue share on transactions, with your own markup on top if you set one. Issuing costs $5,000/mo with $0 per card and the same 1% share. Purchase Agent is $0.15 per request and free while in beta. Enterprise tiers for Vault and Issuing are quote-based and add SSO/SCIM, SLA, a SOC 2 report and a custom contract. Named customers on the page are Orchid (consumer), Almanac (B2B via a Slack agent) and Laso Finance (crypto). Put Agentcard next to Stripe Issuing or Lithic and the difference is intent: those give you card primitives, this gives you card primitives plus a checkout
Behind the Verdict
The pitch is narrow on purpose, and that narrowness is the reason to look. Agentcard is not trying to be a corporate spend card for humans. It is trying to answer a single question: how does an AI agent pay for something without handing it a credential that can be abused? Pick this when your agent already drives a browser and gets stuck at the payment field. The Kernel, Browserbase, Browser Use, Anchor and Playwright attach paths put the card in at the network or session layer, and the customer's own card stays in the vault — so you're not building a PCI scope around the merchant checkout. In practice that's a weekend of work instead of a quarter. Pick this if you'd rather not run a browser stack at all. Purchase Agent is one API call or one MCP buy tool, covers 18+ named merchants, confirms the exact cart, and returns a changed price before paying. At $0.15 a request (free in beta) it's a reasonable way to test whether agentic purchasing has any demand in your product before you staff it. Pass on Agentcard if your buying happens outside Amazon, Walmart, Target, Best Buy, Home Depot, Lowe's, Macy's, Wayfair, Staples, Kohl's, B&H Photo, DoorDash, Good Eggs, Locale and flights — those are the covered merchants, and everything else depends on you supplying your own browser automation. Pass too if procurement requires a completed SOC 2 and PCI DSS review in hand. Both are available through the Enterprise tier, but that's a sales conversation, not a download. The closest alternative depends on what you already own. If you're on Stripe for payments, Stripe Issuing gives you programmable cards but leaves checkout interception, approval UX and the swap-in mechanics to you. Lithic is closer on raw card issuing, with the same gap. Agentcard's actual differentiator is that it
Researching Agentcard? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Agentcard actually fits — and what changes day-one when you adopt it.
Store the customer's existing card in Vault, then have the agent drive a Browserbase CDP session to amazon.com with an Agentcard attached at $23.06. The checkout pauses at the processor, the customer gets a passkey approval link, and the order completes on approval.
Outcome: The agent finishes a real purchase and the customer keeps their own card's points, with the amount locked to the exact cart.
Prefund a company balance, issue a multi-use card with a monthly cap, and let the Slack agent place Staples and Best Buy orders. Purchase Agent handles the checkout on covered merchants with one API call.
Outcome: Company purchases run automatically under a spend ceiling, with order numbers and delivery windows visible for reconciliation.
Hold user funds as USDC on Base, issue a one-time Visa card locked to a single merchant and amount, and let the user's agent pay at checkout.
Outcome: Crypto balances become online spend without the user converting to a bank card or giving the agent a reusable number.
Use Cases
- Let an agent buy Amazon or Walmart items at an exact price with a card scoped to that one purchase.
- Give a corporate Slack agent a card ceiling so it can order supplies without a human filling in a form.
- Store customers' cards in Vault so an agent can buy on their behalf while they keep earning points.
- Automate DoorDash, Good Eggs or Locale grocery orders under a per-order limit set by the user.
- Book flights from an agent with a one-time card that can't be reused after checkout.
- Issue multi-use cards with monthly caps so an agent can pay for recurring subscriptions.
- Pay developer rewards in Anthropic or OpenAI tokens instead of points via Tokenback.
- Convert a crypto balance held as USDC on Base into a spendable card for a user's agent.
Models Under the Hood
as of 2026-10-17
Limitations
- Issuing is priced at a flat $5,000/mo plus a 1% revenue share on every transaction, with no cheaper published tier for smaller teams; above the Vault Free cap of 5,000 users you move to a contact-sales Enterprise plan.
- Cards are store- and amount-locked at issuance, so each intended purchase needs its own card, and the Purchase Agent covers the named merchant list (Amazon, Walmart, Target, DoorDash and others) rather than any store.
- Auto-approval and 12-digit network cards have edge cases — payments at merchants whose checkout encrypts the card from their servers are still refused per the changelog.
- Security attestation (SOC 2 report, questionnaires, pen-test summary) is offered as part of the Enterprise security review rather than published up front.
as of 2026-10-04
Verification history
We have re-verified Agentcard 10 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 10 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Agentcard tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Vault Free
$0/mo
Vault Enterprise
Custom
Ideal for
Established app past 5,000 users, or a regulated buyer needing SSO, a signed SLA and a published SOC 2 report before going live.
What this tier adds
Adds unlimited users, SSO and SCIM, a named support contact with a shared Slack channel, written SLA, custom approval rules and a security review package.
Issuing
$5,000/mo
Ideal for
Company with real agent checkout volume that needs one-time and multi-use cards issued from a prefunded balance or user wallets.
What this tier adds
Flat $5,000/mo with $0 per card, 1% revenue share back to you, your own markup, Tokenback rewards, USDC on Base funding and production in under a week.
Issuing Enterprise
Custom
Ideal for
High-volume issuer that needs rates that fall with card volume, plus enterprise identity, security and contracting terms.
What this tier adds
Adds volume pricing, dedicated support with SLA, SSO and SCIM, custom limits, SOC 2 security review and a custom contract with DPA.
Purchase Agent
$0.15/request
Ideal for
Team that wants agent purchases on Amazon, Walmart, Target, DoorDash and similar merchants without building or running its own browser automation.
What this tier adds
Starting tier for the managed shopping agent — $0.15 per request (free in beta), one API call or the MCP buy tool, passkey approval paid from the user's Vault card.
Where the pricing makes sense
The company stage and team size where Agentcard's pricing actually pencils out — and where peers do it cheaper.
Free Vault under 5,000 users is genuinely cheap to evaluate, and Purchase Agent at $0.15/request (free in beta) lets you ship one merchant flow without a card-issuing contract. Issuing at $5,000/mo plus 1% revenue share is priced for companies with real agent checkout volume — a serious commitment next to a hobby-scale Stripe Issuing setup, and below enterprise card-issuing contracts that usually require far larger minimums.
Setup time & first value
How long it actually takes to get something useful out of Agentcard — broken out by persona, not the marketing-page minute.
Consumer-app engineers on Vault can store cards and attach Agentcard to a Playwright or Kernel session in an afternoon, and the CLI wizard (`npx agent-cards companies wizard`) is advertised as a 10-minute path into sandbox. Purchase Agent is one API call or the MCP buy tool, and the pricing page claims live in an afternoon with no merchant integrations to build. Issuing in production is quoted as
Switching to or from Agentcard
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From raw Stripe Issuing: keep your issuing keys but move agent purchases onto Agentcard so cards are merchant- and amount-locked and gated by passkey approval.
- →From a shared corporate card: replace the single number with one-time Agentcards scoped to the intended store and amount.
- →From hand-rolled browser checkout: attach Agentcard to your existing Kernel, Browserbase, Browser Use, Anchor or Playwright session instead of typing card numbers yourself.
- →From manual reimbursement: prefund a company balance and route agent purchases through Issuing so charges settle instantly rather than getting expensed.
- →From a points-only rewards program: switch to Tokenback so users earn Anthropic and OpenAI tokens or USDC on Base on every agent purchase.
- ↗To Stripe Issuing: recreate your virtual-card issuance and spend controls directly on Stripe's issuing stack if you need a broader merchant surface.
- ↗To Lithic: rebuild card issuing where you need a different card-program configuration or on-prem-adjacent controls.
- ↗To your own acquirer: keep the browser automation but replace Agentcard's approval gate with your own hosted checkout.
- ↗To a human checkout flow: drop the agent and have customers complete purchases themselves if agent commerce volume never justifies the $5,000/mo Issuing plan.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “Agentcard”, and we withheld 6: 6 could not be judged, because “Agentcard” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Agentcard.
Official links
Featured Head-to-Head Comparisons
Agentcard vs Truleo
Truleo and Agentcard serve entirely different markets. Truleo is a must-have for law enforcement agencies drowning in siloed data and manual casework — it chops report writing from 40 minutes to 7 minutes and auto-surfaces leads. Agentcard is essential for developers building AI agents that need to make purchases safely, with single-use cards and human oversight. Choose based on your domain: policing or agent payments.
Agentcard vs Presto Voice
If you run a QSR chain, Presto Voice is your clear choice for driving revenue through automated drive-thru voice AI. If you're a developer or power user letting AI agents spend money online, Agentcard provides secure, single-use virtual cards with granular controls. They solve completely different problems—Presto Voice boosts restaurant revenue, Agentcard enables autonomous agent purchasing without exposing your real card.
Agentcard vs Bitsgap
If you need an AI agent to safely spend money online, Agentcard is the obvious choice—it gives each task a single-use card and requires your approval. If you want to automate crypto trading, Bitsgap offers a mature suite of bots and demo mode. They solve completely different problems: agent payments vs. trading automation. Pick the one that matches your use case.
Popular in Agent Memory & Runtimes
Frequently Asked Questions
Best-of guides
Topics
Used Agentcard? Help shape our editorial sentiment research.