pay per call, no key, settle on base.

what farout is

An x402 pay-per-call LLM gateway with true-up (usage-based) billing. Your agent POSTs a chat completion, the server answers HTTP 402 with a USD spending cap, your client pays USDC on Base, retries, and gets the model response. No API key, no signup, no subscription. Settlement is on-chain and happens after the answer — charged by actual token usage, never more than the cap.

• base url: https://farouter.tech
• paid endpoint: POST /v1/chat/completions ([OI]-compatible)
• anthropic alias: POST /v1/messages
• catalog + live prices: GET /v1/models
• manifest: GET /.well-known/x402

the payment flow

1. POST your request without any Authorization header.
2. Read the 402. The body carries a quote (your spending cap: USD, token estimate, per-1M prices). The payment-required response header carries the base64 x402 v2 requirements (scheme upto).
3. Build the payment with any x402 client and retry with the payment header. Official SDKs send PAYMENT-SIGNATURE; FarOut also accepts PAYMENT and X-PAYMENT.
4. On success: 200 plus a payment-response header with the on-chain settlement receipt (tx hash). The settled amount is based on actual usage (true-up), not the cap.

The cap is computed from your request: estimated input tokens plus your max_tokens, at the model's per-1M price. After the model answers, settlement charges the actual usage (input + cached reads at 10% + output) — floored at $0.001, capped at $3.00. Unused budget is never taken: no refunds needed, the leftover simply isn't charged. Verify any tx on basescan.org. Failed upstream calls are cancelled automatically and never charged.

quick start

import { wrapFetchWithPayment, x402Client } from "@x402/fetch";
import { UptoEvmScheme } from "@x402/evm/upto/client";
import { toClientEvmSigner } from "@x402/evm";
import { privateKeyToAccount } from "viem/accounts";

const account = privateKeyToAccount(process.env.BUYER_PK);
const signer = toClientEvmSigner(account);
const client = new x402Client()
  .register("eip155:8453", new UptoEvmScheme(signer));
const fetchWithPay = wrapFetchWithPayment(fetch, client);

const r = await fetchWithPay(
  "https://farouter.tech/v1/chat/completions",
  {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      model: "glm-5.3",
      messages: [{ role: "user", content: "hi" }],
      max_tokens: 50,
    }),
  });
const answer = await r.json();

Requires npm install @x402/fetch @x402/evm viem. The buyer wallet needs a little USDC on Base. Signing is gasless (Permit2); gas ETH is only needed for the one-time Permit2 approval setup. Raw machine-readable versions: /docs/agents.md and /docs/skill.md.

request fields

response codes

cost model (true-up)

• the 402 quote is a spending cap, not the final price. You authorize the cap with one gasless Permit2 signature (x402 upto scheme).
• after the model answers, settlement charges actual usage: real input tokens + cached reads at 10% + real output tokens. Always <= the cap.
• unused budget is simply never taken — no refunds needed. Check any tx on basescan.org.
• minimum charge per successful call: $0.001. Cap: $3.00 per call (the live value is in every 402 quote).
• probe free: POST without paying to read quote.usd (the cap) before committing.
• failed calls are cancelled server-side; you are never charged for a 502.
• free gas sponsorship: prove you're a paying customer (one settled call, or proofTx — tx hash of any USDC transfer from your wallet to the FarOut wallet) and your wallet holds exactly 0 ETH, then POST your wallet to /api/sponsor-allowance and FarOut sends 0.000002 ETH for your one-time Permit2 approvals (upto setup). One drop per wallet.

model choice

• cheapest: glm-5.3-flash, deepseek-v4-flash, gpt-5.6-luna
• balanced: glm-5.3, glm-5.2, kimi-k3
• flagship: gpt-5.5, gpt-5.6-sol
• live list and prices: GET /v1/models; health: GET /api/status. Ids carry no provider prefix.

rules of thumb

• quote first when cost matters. The unpaid probe is free — quote.usd is your worst case.
• your real cost is almost always below the quote: true-up settles at actual usage.
• one payment header per request. "payment already attempted" from the SDK means rebuild the request with a fresh payload.
• rate_limited in health data is an upstream bid squeeze, not an outage. Retry or switch models.
• keep max_tokens reasonable; it sets the cap you authorize (actual usage is what gets settled, but the cap must cover it).

verification

• unpaid POST returns 402 with a quote in the body (scheme upto).
• paid POST returns 200 with model content.
• the payment-response header contains a tx hash that resolves on basescan.org, and the transferred amount is <= the quoted cap (true-up billing).
• GET /v1/models lists the models you use.