Every agent-payable API.

Turn a text into a structured outline: title and sections with key points, faithful to the text.

Translate text into any language, with the source language detected and tone and formatting kept.

Sentiment of a text: positive, negative, neutral or mixed, a score from -1 to 1, the emotions present and a one-line reason.

Detect the language of a text: ISO code, language name, script, and confidence.

Read back values the paying wallet saved with /paid/memory/set: one key, or list your keys by prefix (keys, sizes and expiry, newest first). Another wallet can never see them.

Line-by-line difference between two texts (up to 3,000 lines each): added, removed and unchanged lines in order, with counts.

Remove personal information from a text before your agent stores or shares it: emails, phone numbers, names, street addresses, IDs and card numbers are replaced with labels.

Text to speech: turn up to 2,000 characters into natural-sounding speech (MeloTTS), returned as base64 MP3. Languages: English, Spanish, French, Chinese, Japanese, Korean.

Pull structured fields out of free text: name the fields you want and get JSON back (missing fields come back null). For invoices, emails, listings, bios and more.

Chat completion from an open AI model (Llama 3.3 70B by default; Llama 3.1 8B for speed), no account or API key: send messages (or a prompt), get the model's reply. Up to 24,000 characters in and 1,024 tokens out per call.

Rewrite a text in a requested style (clearer, shorter, formal, friendly, plain English…) keeping its meaning.

Classify a text into topic categories: from your own list of labels, or open topics if none given, with confidence.

Named entities in a text: people, organisations, places, dates, money amounts, products and other notable items, each with its type and how many times it appears.

Extract the key words and phrases from a text, most important first, with a one-line topic.

Read what arrived at your webhook inbox (from /paid/inbox/create): method, headers (cookies and auth removed), query and body of each delivery, newest last. Pass after= the last id you saw to get only new ones; delete=1 clears what you read.

Get a webhook URL for an agent with no server: anything sent to it (any method, up to 64 KB, 2,000 a day) is kept for 30 days and read back with /paid/inbox/read. Only the paying wallet can read it. Use it for payment notices, job callbacks or alerts.

Status of your scheduled callbacks: next call, calls left, last HTTP status, timing and error. Pass id for one schedule, or nothing to list all of yours; cancel=1 stops it.

Proof that a hash was stamped: when we received it, the Merkle path from its leaf to the root, and the Base network transaction that recorded the root, with step-by-step instructions to check it yourself without trusting us.

Live US-dollar prices for up to 25 crypto assets in one call, by ticker (BTC, SOL…), feed id, or chain:address (e.g. base:0x…), with symbol, confidence and timestamp.

Perpetual futures data from a major on-chain derivatives exchange: hourly funding rate (and annualised), open interest, mark and oracle price, 24h volume, for one or all markets.

Crypto market Fear & Greed index: today's score (0-100) and classification, plus up to 30 days of history for trend.

Historical US-dollar price of up to 10 crypto assets on a given date (or Unix time), by ticker or chain:address.

Position-size and risk calculator for trading agents: from account size, risk % per trade, entry and stop (optional target and leverage) get the position size, notional, money at risk, reward-to-risk ratio, break-even win rate and an estimated liquidation price. Pure arithmetic, not financial advice.

Live prediction-market odds: the most traded open markets (or those matching your search) with each outcome's implied probability, volume, liquidity and end date.

Official daily foreign-exchange reference rates for about 30 currencies, in any base currency, with optional amount conversion.

Technical indicators for a crypto asset from daily prices: SMA 20/50/200, EMA 12/26, MACD, RSI 14, 30-day volatility and drawdown, with a plain trend label. Analysis aid, not financial advice.

Top stablecoins by circulating supply: price, deviation from peg, peg type, supply in US dollars and number of chains; flags any that have lost their peg.

Large US-dollar stablecoin transfers on the Base network in recent blocks (default 100,000 dollars and up): from, to, amount, block and transaction, biggest first.

Before your agent pays or accepts funds: check up to 20 wallet addresses against the US Treasury OFAC sanctions list (SDN digital-currency addresses, EVM format, refreshed daily). Returns a match flag per address and the list date. Screening aid, not legal advice.

Profiles for up to 10 Base network wallets in one call: wallet or contract, native and US-dollar stablecoin balances, and transaction count for each.

Security pre-check for any Base network contract before your agent interacts with it (vault, router, token, bridge): built-in powers in its code (mint, pause, blacklist, upgrade, withdraw, self-destruct, delegatecall), proxy and who controls it (a single wallet or a multisig contract), whether the source is publicly verified, with red flags, a risk level and a plain-language verdict. Facts, not financial advice.

AI risk verdict for a Base network token contract address: send the address, get on-chain facts (name, symbol, supply, owner and whether ownership is renounced, upgradeable proxy, owner and contract holdings, burned share), red flags, a risk level and a short AI-written verdict. Use before your agent buys, accepts or lists a token. Facts, not financial advice.

Recent event logs of a Base network contract (up to 1,000 blocks back, 200 events), with token Transfer and Approval events decoded (from, to, amount).

Live US-dollar price of any Base network token, read directly from on-chain exchange pools (direct stablecoin pools and routes through wrapped ETH; the best quote wins). Returns price, route and pool fee tier.

Explain or verify a Base network transaction: status, from/to, native value, fee, every token transfer (with symbols and amounts) and a one-line plain summary. Use it to confirm a payment arrived before delivering work.

Which public Base network RPC endpoints are healthy right now: each one's latest block, how far behind the freshest it is, and response time. Pick a fast, in-sync node before you send.

What is this Base network address? Wallet or contract, bytecode size, upgradeable proxy and implementation, token standard (ERC-20, ERC-721, ERC-1155) with name and symbol, owner, nonce and balances.

A Base network block (latest, or by number): hash, time, transaction count, gas used and limit, base fee and fee recipient.

ERC-20 token balance of any Base network wallet (the main US-dollar stablecoin by default): raw units, decimals, symbol and formatted amount.

Current Base network gas: base fee, suggested priority fee and max fee, and the cost in ETH of a plain transfer and of a typical token transfer.

Live balances of a Base wallet: native ETH and US-dollar stablecoin, plus its transaction count.

Exact finance maths: loan repayment and total interest, compound growth with regular contributions, NPV and IRR of cash flows, and ROI with annualised return.

Cheapest real end-to-end test of your pay-per-call (HTTP 402) client: pay US$0.002, get back proof of what was paid (payer, amount, network, timestamp) and the request you sent. Use it to check a wallet, a client library or a facilitator before buying something bigger.

Explain a cron expression in plain English and list its next run times in any time zone (5-field cron, names like MON/JAN, ranges, steps and @daily-style macros).

Fresh random identifiers from a secure generator: UUID v4 (up to 50) plus random hex bytes.

Convert CSV (or TSV) to JSON rows with the header row as keys; numbers and true/false are typed, quoted fields and embedded newlines handled. Up to 50,000 rows.

Repair broken JSON (typical of model output) into valid JSON: removes code fences, comments and stray text, fixes single quotes, unquoted keys, trailing commas, Python literals (True/None) and truncated output. Optionally checks the result against a JSON Schema.

Public holidays for a country and year, and optionally the number of business days between two dates (weekends and national holidays excluded).

Hash a text with SHA-256, SHA-384, SHA-512 or SHA-1: hex and base64 digests, for checksums, deduplication and content fingerprints.

Cryptographically secure random integers in a range you choose (unbiased), up to 1,000 per call.