Normalize an email address for matching: send an address and get its local part, domain and a lowercase comparison form. Syntax only; it does not verify the mailbox. Pair with email-auth-check for the domain. $0.001 USDC per call on Base.

Marine weather and sea-state forecast for any coastal coordinate. Returns significant wave height, wave direction and period, wind waves, swell (height, direction, period) and ocean current velocity and direction. Oceanographic conditions for shipping, offshore, maritime safety, fishing, surfing and coastal operations.

Count weekdays (Mon-Fri) between two dates, inclusive of both endpoints.

Prüft am Objekt pixel den Befund Übermittlung vor Einwilligung.

DeFi yield and APY data for AI agents and yield optimizers. Thousands of liquidity and lending pools across chains and protocols, filterable by chain, project, token symbol, minimum TVL and stablecoin-only, sorted by APY. Each pool returns chain, project, symbol, TVL and APY. Sourced from DefiLlama yields.

Make a URL slug: send a title or phrase and get a lowercase, hyphenated, URL-safe slug. Use it for file names, page paths or IDs. $0.001 USDC per call on Base.

River discharge and flood-risk forecast (Copernicus GloFAS, 5 km grid) for the nearest river to any lat/lon. Returns current river discharge in cubic meters per second (m3/s) plus mean, median, max, min and 25/75 percentiles. Hydrology and natural-disaster data for flood forecasting, insurance, agriculture and river logistics.

chainpeek (Tanod) account nonce: the transaction count of an Ethereum, Base or Robinhood Chain address (latest or pending). Use when an agent must read an address's transaction count to set the nonce of its next transaction.

chainpeek (Tanod) contract storage slot read (eth_getStorageAt) on Ethereum, Base or Robinhood Chain. Use when an agent must read a raw storage slot of a contract, for example a proxy implementation slot.

Base64-encode text: send UTF-8 text and get its Base64 string. Use it to build headers, data URLs or API payloads that need Base64. $0.001 USDC per call on Base.

Base64-decode safely: send a Base64 string and get the UTF-8 text, or a clear error when the input is not valid Base64. $0.001 USDC per call on Base.

Prüft am Objekt netzwerkanfrage den Befund Übermittlung vor Einwilligung.

Prüft am Objekt eingabefeld den Befund Übermittlung vor Einwilligung.

Prüft am Objekt formular den Befund Übermittlung vor Einwilligung.

Prüft am Objekt iframe den Befund Übermittlung vor Einwilligung.

utilpeek (Tanod) language detection: the language of a text (75 languages, offline). Use when an agent must identify which language a piece of text is written in.

Get one canonical form for a URL: send an http(s) URL and get one deterministic normalized form. Use it to dedupe links or cache keys. $0.001 USDC per call on Base.

Liquidation pressure feed: 5-second buckets per symbol of long and short USD liquidated across venues, the dominant side and pressure against the symbol's own baseline, from a gap-free cursor. Use when you need to answer: Which side is being liquidated across venues right now, and how hard compared with normal? One call per 5 s, 10 s on a pass. Heavy pollers: $1 day pass, /api/keys/buy-day-pass.

Sanctions screening for compliance, AML, KYC and KYB agents: use before onboarding, paying or contracting with a person or organization. Fuzzy-matches the name against the official OFAC SDN list and the UN Security Council consolidated list, aliases included, and returns scored matches with list, sanctions program, entity type and source reference. Diacritic- and case-insensitive, tunable min_score (default 0.6). One keyless USDC call.

OpenAI-compatible chat completion on 22 models: name one in `model` (meta-llama/Meta-Llama-3.1-8B-Instruct, deepseek-ai/DeepSeek-V4-Flash, deepseek-v4-flash, openai/gpt-oss-120b, openai-gpt-oss-120b, meta-llama/Llama-3.3-70B-Instruct and 15 more) and the 402 prices that model per request; served by DeepInfra, Venice AI, NEAR AI Cloud and Taifoon's own GPU (nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16 when no model is named). Messages in, the answer and its token usage out; max_tokens at most 700.