Web & Data

Get details for a Reddit community (subreddit).

Screen a declared subject against independently retrieved official sanctions sources and return traceable evidence

Normalize public text, HTML, JSON or XML documents into source-backed JSON

Read any public web page as clean markdown — rendered, boilerplate stripped, ready for a context window. Use it when your agent needs the content of a URL and has a wallet instead of a scraping stack. Charged ONLY if the fetch succeeds; a failed fetch costs nothing. stealth=true ($0.05) gets past aggressive bot protection. Pair with /v1/web/search to find pages first.

Extract up to ten URLs in one paid call.

Fetch a current third-party aggregated model-pricing snapshot from LiteLLM. Buyers should verify final pricing with each provider.

Search a live public Shopify storefront and return current public products, variants, prices, availability, images, and declared agent-commerce capabilities.

Viewport or full page, optional element selector, dark mode and cookie banner removal. Returns base64 image plus a short lived fetch URL, byte size, load time and the final HTTP status of the page.

Return title, description, OpenGraph card, canonical URL, feeds, favicon and detected technologies for one URL.

Find academic papers in any field - title, authors, year, venue, DOI, citation count, open-access link and abstract - from OpenAlex (~250M works incl. arXiv and PubMed; Crossref answers if OpenAlex is down). Filter by year range and open access; sort by relevance, citations or date. Send {} to search the published example query. A search with no result is not charged. Pay-per-call via x402.

Return the structured offer of one product page: identifiers, price, availability, shipping, variants, ratings, images and seller, each with its source.

AI data lookup: ask any question, get facts + sources.

Fetch one URL and return clean article Markdown without navigation, ads or cookie banners.

Returns computed results and answers to queries processed through a computational knowledge engine.

Returns a list of nearby locations based on geographic search parameters.

Retrieves customer reviews for a specified location.

Provides simplified computational answers and mathematical results for queries.

Turn messy text — or any URL's page content — into structured JSON matching the shape you ask for, so an agent can consume a web page as data instead of prose

General market news with category filtering

Search a governed snapshot of implemented, validated Polymathica PDE and scientific-computing solvers. Returns validation versions, bounded evidence locations, implementation modules and provenance without claiming a fresh solver execution.

Return a bounded, source-labelled set of public-domain Bible passages and neutral reflection prompts for one of six predefined topics.

Return bounded ranked OpenBible.info cross-reference identifiers with displayed text resolved only from BSB or WEB.

Live web search for agents: up to 10 results with title, URL and snippet. Use it when your agent needs to find current pages and has a wallet instead of a search API key. Charged ONLY if the search succeeds. Results pair with /v1/web/read, which returns any of them as clean markdown.

Publicly verifiable randomness from the drand League of Entropy beacon, attested and timestamped. Use it for a fair draw between parties who do not trust each other: a distributed network emits one signed value every 30 seconds and no participant, Soren included, can choose or predict it. Soren refuses to attest a round that does not verify, and you can check the same round yourself at api.drand.sh.