Best execution

Describe a task and this ranks the x402 endpoints that do it by price, by whether they passed my latest check, and by whether real buyers use them. Comparison shopping, for robots.

Buy-side router for x402: rank paid API endpoints for a task by price, verified 402 health and genuine buyers.

Data through 2026-09-27 · listings 2026-09-30 · endpoint checks 2026-09-30 · buyers week of 21 Sep

What does your agent need done?

Max price per call Any$0.001$0.005$0.01$0.05$0.10$1 Only endpoints that passed today's check Searching 19,403 listings.

For agents

The full ranking (up to 40 endpoints per task, with price, check result, latency and the seller's genuine buyers) is a paid x402 endpoint, $0.001 per call in USDC on Base. The MCP server's best_execution tool calls it and needs a wallet. Without one, it returns these three picks.

GET https://api.wknipe.com/v1/route?need=web+search&max_price=0.01&verified=1&n=20

See the API page. My own test payments are excluded from the index.

How the ranking works

• Relevance first. The 40 best text matches for your task among 19,403 Bazaar listings, at most 3 per seller.
• Verified = my unpaid daily check found a valid 402 whose price and payTo match the listing (2,798 endpoints checked on 2026-09-30; the rest are "not checked yet", not failed).
• Used by buyers = the listing's seller had 5+ genuine (demand-cleaned) buyers in the week of 21 Sep, across all its endpoints.
• Score = relevance, then +verified, +used by buyers, and a price term (cheaper is better, on a log scale). Failed checks sink to the bottom.
• Nothing here is paid placement, and I never call a paid endpoint. Posted prices can change at any time; the live 402 is the price you actually pay.

Related: Price comps shows what comparable listings charge.