FAQ

Payment wires: every paid endpoint accepts x402 and MPP (Machine Payments Protocol) on the same 402 - see Paying with x402 and Paying with MPP. Agent402 is the applied layer of Agentic Finance: agents that pay and get paid on their own.

Do I need an account or API key? No. Nothing here has a signup. Payment (USDC over x402 or MPP, or proof-of-work) is the only credential, per call. A prepaid credits key (a402_…) also pays Agent402's own tools with one header; it is a bearer key, not an account. /credits says whether packs are on sale (see Reports and Monitors).

What does it cost? Flat per-call prices starting at $0.001. Most tools are $0.001 to $0.02; premium inference, media, and multi-tool skill packs are priced from the tools they run ($0.003 to $0.101). Every price is published in /api/pricing and quoted exactly in every 402 response. Report products (/v1/research, /v1/dossier, /v1/ticker-pack, /v1/fund, /v1/filing-report, /v1/domain-audit, /v1/recall-report, /v1/insider-report, /v1/token-brief, /v1/token-risk, /v1/linkedin-article) are priced per finished report, $0.60 to $2.00 over x402 or MPP ($0.05 for the deterministic /v1/ipo-report digest), or $2 to $5 by card at /reports - the card price includes payment processing, and an agent paying per call pays the lower tool price for the same report. Monitors at /monitors are the one subscription, $5 per month per target. The LLM gateway's metered tier (POST /v1/metered/chat/completions) is the one route whose 402 quotes each request from its body rather than a flat price.

Can I pay by card? Yes: a finished report at /reports or a monitor at /monitors. A prepaid credits key pays Agent402's own tools (debited only on a successful call, never expires, no fees; /credits says whether packs are on sale). Routes at or above the card minimum (fifty cents) also accept cards over the MPP wire (Stripe stripe/charge) when the operator enables it. Neither a credits key nor a card pays the routes that buy from an outside seller (the route-execute tiers, seller-payability, /api/decide/execute and the flight routes); those take a wallet payment. Identity-bound tools (memory, my-usage) need a wallet payment because the wallet is the identity.

Can I use it without any money? Yes: 150+ pure-CPU tools accept proof-of-work (sub-second of your CPU), and the hosted MCP connector runs the same set free (rate-limited) through catalog.call. See Paying with Compute.

What is x402? An open HTTP payment standard built on the 402 Payment Required status code: the server quotes a price, the client signs a stablecoin payment and retries, and a facilitator settles it on-chain. See Paying with x402.

Which chain/asset? USDC on Base (primary), Solana, Polygon, Arbitrum, Monad, Celo, Avalanche, Sei, Optimism, Arc, Stellar, or Algorand - plus USDG (Global Dollar) on Robinhood Chain. The buyer needs only the stablecoin - gas is sponsored by the facilitator on the EVM chains. Fee handling on the other rails is described per rail in Paying with x402 and Paying with MPP.

Does using this spend my AI tokens? No. Not your tokens, ever: you never hand us a model key, and a model we run is billed to us and priced into the call. Most of the catalog is deterministic code (parsers, hashes, math, a real browser) with no model in its path. The ones that do run a model are named: the /v1 gateway tiers and judgment routes, the report products, Decide (/api/decide), the judgment model the Smart Order Router uses to pick among shortlisted tools, and the image, speech, transcription, embedding and AI-answer tools; each carries modelBacked: true in /api/pricing. Proof-of-work spends your CPU; x402 or MPP spends USDC.

Is my data stored? Tool inputs are processed in memory to compute the response. The memory tools store data by design (wallet-keyed, owner-deletable, with optional TTL). An input is otherwise kept only in the cases the privacy policy lists, such as short-lived response caches, catalog searches that find no tool, decisions bought on /api/decide (30 days), and the feedback you leave on a paid call. Full policy: agent402.tools/privacy.

How do I know the service is honest? The server is fully open source; CI re-tests every non-metered tool against its own documented example before each deploy (the metered ones - LLM tiers, report products, search and the tools that need a third-party key - are excluded so CI never spends upstream); and revenue settles on-chain to published receiving wallets: agent402.base.eth on Base and the other EVM chains, which anyone can audit on Basescan, and every other rail's address in /.well-known/x402.

What if a tool fails after I paid? Over x402 you weren't charged. The paywall runs the handler first and settles afterwards, and it settles only a response with a status below 400. Any 4xx or 5xx cancels settlement, so a failed call takes no money. (A 200 is charged only once settlement succeeds; if settlement itself fails you get a 402, not a bill.) A payment that is final before the call runs (a Tempo push transfer over MPP) is recorded when the call fails and repaid on-chain; look it up by its transaction with GET /api/refunds/lookup?tx=<hash>. A few tools charge a 200 that carries a partial result by design and say so in their description (a code run that hits its time limit, for example). On top of that, anything that can't be served reliably is removed from the catalog rather than left to fail repeatedly. Failure rates are watched by CI and a 15-minute production heartbeat.

Why did a call answer 503 saying today's usage limit is reached? Tools whose upstream is billed per call have a daily call ceiling. Once a tool's upstream reaches it, calls to that tool answer 503 until 00:00 UTC, when the count starts again. A 503 is not settled on x402, so over x402 that answer takes no money; retry after 00:00 UTC or use another tool for the same job (/api/find lists alternatives).

My video request timed out. Do I pay again to get the clip? POST /v1/videos/generations answers 504 when a clip is not ready in time; a 504 is not settled on x402. The generation keeps running, so a wallet payer who sends the same request again unchanged within the window the 504 message names (30 minutes by default) collects that same clip instead of starting a new one, and pays only for the answer that delivers it.

Can I list my own service alongside this, or integrate? Yes, for free: paste your origin on /sell for an immediate probe, call POST /api/index/register directly, or let the hourly auto-discovery from public x402 discovery registries pick it up once it is live and settling. A listed seller is routable by the Smart Order Router and ranked on the leaderboard. To talk about a deeper integration, open an issue.

Can I find tools on other x402 sellers from here? Yes - Agent402 is also an x402 Index + Smart Order Router. POST /api/route ranks tools across every x402 seller it has crawled (auto-discovered from public x402 discovery registries, refreshed hourly) and filters out unhealthy ones. Candidates are shortlisted by how well they match the task, then ordered by crawl health, by distinct payers over the last 30 days, by price, by a discovery listing's curated flag and by slug length, shorter first; a judgment model can then pick the one that does the job from that shortlist, or decline them all (each result lists the order in why.tiebreaks). Browse the live index at /marketplace.

How do I see which x402 sellers are most used? GET /api/leaderboard returns the head of the live on-chain ranking of x402 sellers by Base USDC settled volume (calls served, totalUsd, unique buyers per seller). It is a top-N slice, not the whole board - 25 rows by default, 50 the ceiling with ?top=N, and totalSellers in the response says how many are ranked in all, so a seller missing from your rows may simply rank below them. The pipeline walks every page of the x402 Bazaar discovery listing, queries eth_getLogs on Base USDC for each seller's payTo wallet, filters per-call settlements within the per-call ceiling reported as maxCallUsd ($0.75 by default; larger inbound is funding, not buys), and aggregates. Snapshot refreshes every 4 hours. Agent402 itself is left out by default; add ?include=all to include it. Full details in x402-Leaderboard.

Who runs this? Havok Holdings LLC - a public, contactable maintainer. Contact: [email protected] · @Agent402Tools on X.

Is there an acceptable-use policy? Who is responsible for generated content? Yes - the hosted instance's Terms of Service include a generative-content acceptable-use policy (and the upstream model providers' usage policies apply to /v1 traffic). Outputs are generated by third-party models from the inputs you send: Agent402 doesn't review, own, publish, or retain them, and you are responsible for your inputs and how you use the outputs. Wallets used for prohibited content are blocked before settlement, so they are never charged. Report abuse: [email protected].