Credia Payments gives each AI agent its own wallet for digital dollars (USDC). Your agent can charge for its work, pay for the APIs and tools it uses, and stay inside the budget you set. You get a signed notification for every payment.
Get started with Credia Payments.
Credia Payments
Status: Live
Payment workflows and controls for AI agents.
You build an AI agent that needs to pay for APIs and tools, or get paid for its work.
You run an API, an MCP server or a paid tool, and you want AI agents to pay you per call.
AI agents are starting to do real work for people and for other software: writing reports, reviewing code, calling paid APIs. To be useful in business, they need to get paid for that work and to pay for what they use, without a person approving every transaction.
Credia Payments gives each agent its own wallet. You decide how much it may spend. When money arrives or leaves, your system gets a signed message so it can deliver the work or update its records.
Sign up with your email and a one-time code. Name your agent and its wallet is created on Base.
Choose how much the agent may spend per payment and per period. An agent that only receives can have a spending limit of zero.
Use the REST API, an official SDK, or a ready-made integration for frameworks like LangChain, CrewAI or MCP.
Payments land in the agent's wallet and you get a signed webhook. Outgoing payments go out only within the limits you set.
Real situations teams use it for today, and exactly how each one works.
Your agent writes research reports. You want anyone to be able to order one and pay, without building a checkout or handling cards.
How it works
You start earning without a payments team, and every sale lands in the agent's own wallet.
AI agents are calling your API, but they cannot sign up for an account or enter a card.
How it works
Agents pay per call automatically. Your existing customers keep using your normal sign-up and keys.
You run an MCP server. Some tools are cheap to run, others cost you real money on every call.
How it works
Your expensive tools pay for themselves, without rebuilding your server.
Your agent needs paid data, another API or another agent's service to finish its job.
How it works
The agent completes the task on its own, and you see exactly what it spent and why.
You want your agent to work overnight without a person approving every payment, but you are worried about a bug or a manipulated prompt.
How it works
The worst case is a number you chose in advance.
You want an AI assistant to check balances or make small payments as part of a workflow.
How it works
Your assistant can handle routine payments, inside boundaries you control.
You want to wire up payments and webhooks before any real money moves.
How it works
You ship payment flows with confidence, and the code does not change between test and live.
You run several agents and need clean records for accounting or a client report.
How it works
Each agent has its own clear history, instead of one wallet everyone shares.
The same path every time, from the payer's wallet to your code.
From any wallet on Base, through your hosted pay link, agent page or a 402 challenge.
The transfer is confirmed on the network, usually within seconds.
The payment appears in the agent's history with the amount, payer and time.
A payment.received event, signed with HMAC-SHA256 so you can verify it came from us.
Your agent releases the report, unlocks the tool or serves the API call.
A payment shows that value moved. Your code decides what to deliver in return.
Send USDC from an agent wallet with the Python SDK. The same call is available over REST and in the other official SDKs.
from agentcredia import Client
client = Client(api_key="sk_test_...") # sk_test_ = testnet, sk_live_ = mainnet
tx = client.payments.create(body={
"agentId": "agt_123",
"to": "0xRecipient",
"amountWei": "10000", # USDC base units (6 decimals): 0.01 USDC
})Ready-made integrations for popular agent frameworks, or plain REST for anything else.
Review supported wallet connections, networks and currencies.
Explore documented payment endpoints, payment events and integration contracts.
Understand which limits are enforced, where they apply and how operators manage them.
Review the records available for reconciliation and investigation.
Spending limits are checked by the AgentCredia API on every payment sent through it. That does not mean every transaction through an external wallet is governed by the same controls. Review the documented authorization and signing model before delegating access.
Every agent starts on Free. Upgrade only the agents that earn. Receiving and funding are free on every plan; a small fee applies when money leaves an agent's wallet.
No. You sign up with an email address, and the dashboard creates the agent's wallet for you. USDC is a digital dollar, so amounts are shown in dollars. Your code talks to a normal REST API.
Yes, any wallet that holds USDC on Base can pay, including Coinbase Smart Wallet, which signs up with a passkey and no seed phrase. If you must serve buyers who will not use a wallet, keep a separate card checkout for them.
USDC on Base. Base Sepolia is the testnet for building without real money. See the payment documentation for current details.
Yes. Test API keys (sk_test_) run against Base Sepolia, a test network where the USDC has no value. Switch to a live key (sk_live_) when you are ready.
Spending limits. Each agent has a per-payment cap and a per-period allowance that you set in the dashboard. The API checks them before it accepts any outgoing payment, and the agent's own code cannot change them.
Read the signing and custody model in our security documentation before delegating access.
AgentCredia sends a signed payment.received webhook to your server when the payment settles on Base, and the payment appears in your dashboard with the payer's address, amount and time.
No. A payment receipt shows that value moved, not that a task was delivered. Your own code decides what to deliver once it sees the payment.
Establish an agent's identity and its connection to an operator.
Explore IdentityDefine how an agent authenticates and what authority it can use.
Explore AuthInspect specific claims and the evidence behind them.
Explore VerifyUnderstand activity and outcomes in context over time.
Explore ReputationSee if the traffic in your app is human, AI-assisted or agent traffic.
Explore Proof-of-ActorCreate an agent wallet in minutes, set its limits and connect your code. Free to start.