APIs were designed for humans. Not directly—you write code so a human doesn't have to click buttons—but the underlying assumption has always been that a human is in the loop. A human signs up, generates a key, monitors usage, renews billing. The API is a tool; the human is the operator.
That assumption is shifting. AI agents are increasingly the consumers of APIs, and they have different requirements than humans do. They don't want to log into dashboards. They can't manage a credentials file. They need to discover, evaluate, and pay for services autonomously.
We built Emailyze for developers. Now we are building it for their agents.
Three things, each addressing a different part of agent-native API access:
The [x402 protocol](https://x402.org) is an open extension to HTTP that enables per-request micropayments. An agent sends a request, receives a 402 Payment Required response with payment details, settles $0.001 USDC on Base, and retries—this time receiving the result.
No API key. No account. No billing portal.
GET https://api.emailyze.xyz/v1/x402/check/?email=test@mailnull.com
→ 402: pay $0.001 USDC on Base
→ 200: {"is_disposable": true, "risk_score": 95, ...}An agent that carries an EVM wallet can call Emailyze independently—discovering the endpoint, understanding the price, and authorizing payment—without any human involvement.
The [Model Context Protocol](https://modelcontextprotocol.io) is Anthropic's standard for giving AI models access to external tools. With the Emailyze MCP server configured in Claude Desktop, Claude gains two new tools: check_email and check_emails_batch.
Ask Claude "Is this email safe?", and it calls the Emailyze API, interprets the result, and answers in context. No code, no dashboard, no integration work beyond a config file.
Installation and configuration details are in the [MCP guide](/guides/mcp-guide).
The llms.txt standard is a lightweight convention—similar to robots.txt—that tells AI models how to interact with a website. Our file at https://emailyze.xyz/llms.txt describes our API, endpoints, and x402 capabilities in a format optimized for LLM consumption. Agents crawling for available tools find everything they need without parsing HTML documentation.
The old workflow for using an email verification API:
1. A developer finds the service 2. Signs up and creates an account 3. Generates an API key 4. Configures the key in their application 5. Monitors usage in the dashboard 6. Renews or upgrades billing when limits are hit
Every step requires a human. An automated agent building an autonomous workflow cannot do any of this.
The new workflow with x402:
1. Agent discovers https://api.emailyze.xyz/.well-known/x402.json 2. Agent reads the manifest: endpoint URL, price ($0.001 USDC), network (Base) 3. Agent evaluates: is email verification worth $0.001 for this task? Yes. 4. Agent makes the request, pays on-chain, receives the result.
No human involved. The agent decides, pays, and acts.
The most interesting property of x402 is not the payment mechanism—it is the decision-making it enables. When an agent can see the price of an operation and autonomously authorize it, it can reason about costs. "Should I verify this email before processing this order? It costs $0.001. The order is worth $80. Yes."
That kind of cost-aware reasoning is impossible when credentials are provisioned by a human upfront. x402 makes pricing legible to agents in the same way that HTTP status codes make server state legible to clients.
Imagine an AI agent managing an e-commerce store's order review queue. When a new order comes in from an unknown customer, the agent:
1. Checks the email via Emailyze x402 — $0.001, paid autonomously 2. If is_disposable: true and risk_score > 80, flags the order for manual review 3. If provider_type: personal and risk_score < 30, approves automatically
No developer wrote a custom integration. No human configured a billing account. The agent discovered the endpoint, paid for the checks it needed, and made decisions based on the results. The only setup was funding a wallet.
We are listed on [x402.org/ecosystem](https://x402.org/ecosystem) alongside other x402-compatible APIs. As the protocol gains adoption and more agent frameworks add native x402 support, the setup will get simpler still.
We will continue adding agent-native features: structured output formats optimized for LLM consumption, capability discovery endpoints, and usage reports accessible via API rather than dashboard.
GET https://api.emailyze.xyz/v1/x402/check/?email={email} — $0.001 USDC/check on Basehttps://api.emailyze.xyz/.well-known/x402.jsonIf you are building AI agents that need to validate email addresses—at signup, during order processing, or anywhere else—these tools are built for you.