1gateway

Use when an agent, x402 test or CI job needs Base Sepolia gas now: one x402 call, no faucet login or mainnet history. Input: to (recipient address); sends 0.1 ETH (1 ETH: /v1/testnet/base-sepolia-eth/1). Output: JSON txHash (confirmed before settling), explorerUrl, amountEth, from, to. Daily limits (UTC): 5 ETH per address (both sizes), 20 overall; checked before payment. Test ETH has no monetary

Use when contract deploys, x402 tests or CI jobs need more Base Sepolia gas than a faucet gives: one x402 call, no faucet login or mainnet history. Input: to (recipient address); sends 1 Base Sepolia ETH. Output: JSON txHash (confirmed before settling), explorerUrl, amountEth, from, to. Daily limits (UTC): 5 ETH per address (both sizes), 20 overall; checked before payment. Test ETH has no monetary

Use when an agent needs to search and read in one call. Input: q, optional max (1-10, default 5), time_range (day, week, month, year). Output: JSON results (title, URL, date) and, for the top 3, the page converted to Markdown (up to 4,000 characters each; a page that fails is marked, the rest still ship). Provider failure is not charged. Example: ?q=EU+AI+Act+timeline&max=5

Use when an agent must vet a South Korean supplier or counterparty before paying, contracting or onboarding it. Input: one of business_number, dart_corp_code, stock_code or q (exact company name). Output: JSON summary (riskLevel, riskFlags such as closed/suspended or sanctions match, per-section status) plus NTS status, DART profile, 12-month KONEPS award summary and OFAC/EU/UK screening of the En

Use when an agent needs a Korean corporate disclosure in English. Input: rcept_no (14-digit DART receipt number from /v1/kr/filings). Output: JSON with headline, what_happened, key_figures (amounts, ratios, share counts with units), parties, key_dates, why_it_matters, confidence, original Korean title and DART link. Machine-generated paraphrase, not the official text. Unknown rcept_no or model out

Use when an agent needs a sourced answer to a question from the live web in one call. Input: q (question), optional time_range (day, week, month, year). Output: JSON answer (English, inline [n] citations), citations [{url, title, quote}], confidence, notes, sources read. Searches, reads the top 4 pages itself and answers only from them. Provider or model failure is not charged. Example: ?q=what+is

Use when an agent needs what is happening on a topic right now, in English, including Korean and Asian news. Input: topic, optional time_range (day, week default, month, year), max (1-10 headlines, default 10). Output: JSON overview, clusters [{headline, summary, sources}] with duplicates merged, and sources (title, url, outlet, publishedAt). Headlines and links only. Failures not charged. Example

Use when an agent needs specific fields from a web page as JSON. Input: url, schema (URL-encoded JSON Schema, object root, max 30 properties, depth 3, 3,000 chars). Output: JSON data validated against the schema (null when not on the page), title, finalUrl. SSRF-safe; noai pages refused; bad schema refused before payment. Fixed schema: /v1/docs/extract. Example: ?url=https://example.com/&schema={"

Use when an agent must grade a website and say what to fix. Input: url or bare domain. Output: JSON score 0-100, grade A-F, per-category scores (security, https, seo, mail, ai), findings [{id, severity, category, title, fix}] and top fixes. Rule-based from one live fetch plus DNS; raw data: /v1/web/site-profile. Example: ?url=stripe.com

Use when an agent tracks South Korean government policy, regulation or draft legislation in English. Input: optional q (keyword, Korean or English), date (YYYY-MM-DD KST), limit (1-20, default 5). Output: JSON items with Korean title, official link, published time and English brief {title_en, summary_en, agency_en, category, key_points}. Machine translation; check the official page. Example: ?q=AI

Use when an agent is about to email or sign up a contact. Input: email. Output: JSON verdict (deliverable, risky, undeliverable) with reasons, MX hosts, null MX, SPF and DMARC policy, disposable domain, role account, free provider. No SMTP probe, so the mailbox itself is not verified. Malformed address refused before payment. Example: ?email=info@example.com

Use when an agent must see how a page looks (layout, charts, visual QA, previews), not just its text. Input: url, width (320-1920, default 1280), full_page=1 for the whole page, format png or jpeg (default). Output: JSON imageBase64 (decode to the image), contentType, bytes, viewport. Renders JavaScript. Daily cap or bad URL refused before payment. Example: ?url=https://example.com/

Use when an agent needs a finished English due-diligence brief on a South Korean company. Input: one of business_number, dart_corp_code, stock_code or q (exact name); format=json|markdown; news=true|false. Output: JSON with executiveSummary, businessOverview, financialHighlights (each figure source-checked), procurementActivity, recentEvents (DART, 90 days), riskAssessment, sanctionsResult, dataGa

Use when an agent has several URLs (search results, citations) and needs them all as Markdown at once. Input: urls (1-5 public URLs separated by spaces), optional max_chars per URL (500-20000), flags=1. Output: JSON results[] in input order, each status ok (markdown, title, finalUrl) or error (code, message). Charged once if at least one page is read. Example: ?urls=https://example.com/ https://ex

Use when checking what changed at a place between two dates (construction, deforestation, flooding, crops). Input: lat, lon, radius_m (200-10000), before and after dates (20+ days apart). Output: JSON per date: clearest scene, area cloud %, NDVI mean/median, true color, NDVI and false color PNG URLs; NDVI change and reliability. No scene: 404, not charged. Example: ?lat=37.5665&lon=126.978&before=

Use when tracking crops, greenery or land change month by month. Input: lat, lon, radius_m (200-10000), months (1-12), optional end date. Output: JSON series per month: clearest scene, area cloud %, NDVI mean/median, status (ok, cloudy, no_scene), plus min/max and first-to-last change. Numbers only, no images. No scene at all: 404, not charged. Example: ?lat=36.35&lon=127.38&months=6

Use when computing due dates, SLAs, settlement or delivery dates. Input: start date, then days (add -260..260 business days) or end (count business days between), optional country KR, US, GB, DE, FR or JP for public holidays (omit = weekends only). Output: JSON result date or business/calendar day counts, holidays skipped on the way, holiday scope. Example: ?start=2026-12-22&days=5&country=US

Use when an agent judges whether a perp's funding is unusually high or low, or tracks open-interest build-up. Input: coin (e.g. BTC), hours (24-720, default 168). Output: current 1h funding and APR, 24h/7d/30d average APR, current percentile and z-score vs 7d and 30d, share of positive hours, downsampled series, open interest now and change 1h/24h/7d (from our 5-minute snapshots). Example: ?coin=B

Use when hunting crowded perps: which markets have funding far from their own normal and where open interest is jumping. Input: optional limit (1-50, default 10), min_oi_usd (default 1,000,000). Output: as_of, fundingExtremes ranked by |z| (current 1h funding vs the coin's last 7 days), oiMovers24h and oiMovers1h (open-interest % change), each with funding APR and OI USD. Example: ?limit=10

Use when an agent must hand over a document as a PDF (report, invoice, summary, receipt). Input: GET ?url= for a public page, or POST JSON {html | markdown | url, format A4|Letter, landscape, margin_mm}. Markdown gets a clean stylesheet with Korean fonts. Output: JSON fileBase64 (decode to the PDF), bytes, page settings. Headless Chrome on Cloudflare. Bad input or daily cap refused before payment.

Use when an agent must turn its own HTML or Markdown into an image (social card, chart, table snapshot, preview) or capture a page as PNG. Input: GET ?url=, or POST JSON {html | markdown | url, width 320-1920, full_page}. Korean fonts included. Output: JSON fileBase64 (decode to the PNG), bytes, viewport. Bad input or daily cap refused before payment. Example: ?url=https://example.com/

Use when an agent must read text from a scan, photo or screenshot: scanned PDFs (embedded page images) or PNG/JPEG/WebP up to 5 MB, Korean and English, tables kept as Markdown. Up to 3 pages per call. Input: url via GET, or POST a file (multipart 'file', raw body, or JSON contentBase64). Output: JSON markdown, pages with confidence, warnings. PDFs with selectable text: use /v1/convert/markdown. Ex

Use when an agent or seller must confirm a payment before delivering: checks a Base, Base Sepolia or Solana transaction for a USDC transfer to pay_to of at least amount. Input: tx (hash or signature), pay_to, network (default eip155:8453), optional amount in USDC. Output: JSON paid true/false, verdict, received, payers, block, time, confirmations. Bad formats refused before payment. Example: ?tx=0

Use when an agent must screen many wallets at once (payouts, counterparty lists) against OFAC SDN wallet addresses. Input: addresses (1-50, comma-separated, mixed chains) or POST JSON {addresses:[...]}. Output: per address listed yes/no, detected chains, SDN entity, programs; summary counts; list date. Unknown formats refused before payment. Not a legal determination. Example: ?addresses=0x098B716

Use when an agent must check a transaction before accepting or relaying funds: screens its direct parties (sender, recipient, contracts, token Transfer parties on Base; accounts and token owners on Solana) against OFAC SDN wallet addresses. Input: tx, network (eip155:8453 default, or solana). Output: verdict, parties, roles, listed yes/no, SDN entity, programs, tx status, list date. Bad formats re

Use when an agent must tell a user what a transaction did. Input: tx (Base hash or Solana signature), network=eip155:8453|solana (optional, inferred). Output: explanation (1-3 sentences; every number checked against the decoded facts, template fallback), facts: status, time, fee, from/to, called function (curated selector table), ERC-20 transfers, approvals, inferred swap (sent/received), Solana p

Use when you sell an API over x402 and want it to pass Bazaar/CDP listing checks. Input: url (https x402 resource), method GET or POST. Output: JSON score 0-100, grade, per-check pass/warn/fail with fix hints (x402 v2 quote, accepts fields, USDC asset, payTo format and OFAC list, description quality, Bazaar schemas, header size, CORS, latency, openapi/llms.txt, ERC-8004) plus our measured uptime.

Use when an agent is about to pay an unknown x402 endpoint and needs a cheap safety gate. Input: url (https x402 resource), method, optional max_price_usd. Output: JSON verdict ALLOW/WARN/DENY with reasons: live quote now, price vs similar routes and your budget, payTo on the OFAC list, price change vs our last measurement, our 7-day uptime/p95, ERC-8004 registration. Never pays; cached 60 s. Exam

Use when an agent needs one English brief on a token before trading or listing it. Input: token (base:0x… or sol:<mint>), format=json|markdown. Output: executiveSummary, riskAssessment (signals, owner/proxy or mint/freeze authority, holders), liquidity (price, depth, $1k/$10k buy impact), perpMarket (Hyperliquid, symbol match), sanctions (OFAC wallets), dataGaps. AI-written from chain facts; not a

Use when you need to convert an amount between currencies (USD↔KRW, EUR→JPY, ...) at the official ECB reference rate, today or on a past date. Input: from, to (1–10 comma-separated), amount (default 1), optional date (YYYY-MM-DD since 1999-01-04; weekends use the prior fixing). Output: JSON per target rate, inverse rate and converted result, the fixing date used. Source: ECB statistics; reference

Use when you need a historical exchange-rate series or a period summary (e.g. USD/KRW over the last quarter). Input: symbols (1–10), base (default EUR), start and end (YYYY-MM-DD since 1999-01-04; default last 30 days; max 366 days). Output: JSON daily series of ECB fixings, and per currency first/last, % change, min/max with dates, average and daily volatility. Source: ECB statistics; reference o

Use when an agent must compare lending yields on Base. Input: optional protocol (aave-v3,compound-v3,morpho,moonwell) and asset (USDC,WETH,cbBTC,wstETH,cbETH,EURC). Output: JSON per market supply/borrow APY, utilization, supplied/borrowed amount and TVL in USD, sorted by supply APY, with block. Computed from contract state; base rates, rewards excluded; not advice. Example: ?asset=USDC

Use when an agent needs an image from a text prompt. Input: prompt (max 1000 chars), optional steps (1-8, default 4). Output: JSON with a 1024x1024 JPEG (base64), steps, model, safety verdicts. Prompt screened by blocklist + Llama Guard 3, image by a vision check; sexual, minor, deepfake or gore requests refused 422, not charged. Not stored. Example: ?prompt=isometric%20illustration%20of%20a%20tin

Use when an agent needs spoken audio from short text (voice reply, narration, accessibility). Input: text (max 500 chars), lang (en, ko, ja, zh, es, fr; default en). Output: JSON with an MP3 (base64), voice, model, safety verdict. Fixed built-in voices, no voice cloning. Text screened by Llama Guard 3; unsafe 422, not charged. Not stored. Example: ?text=Hello%20from%201gateway&lang=en

Use when an agent wants a quick read on a Base ERC-20 before trading: the full token-risk scan plus a short AI verdict. Input: address. Output: JSON verdict (level = deterministic riskLevel, headline, top reasons citing fired signal ids, what to watch) + all risk signals, simulation, notChecked. Signals, not advice; not a safe verdict. Bad input/EOA: 400, not charged. Example: ?address=0x4ed4E8628