ApiClay

ApiClay is an API marketplace and publishing platform for turning existing data or services into documented endpoints. One side of the product helps developers discover an API, inspect its contract, test it in a playground, and copy integration code. The other side lets a publisher connect a database, spreadsheet, file, live website, OpenAPI service, or custom HTTP source and expose selected data through ApiClay with access rules, usage limits, trials, plans, and per-call pricing.

For buyers, the catalog is meant to reduce the work between finding a data provider and making a reliable request. Each API can carry reference documentation, a playground, and its own pricing information. Buyers use an ApiClay key, while upstream credentials stay on the server. Search and schema inspection are free. A live call follows the chosen listing's trial, plan, and usage rules, and a wallet ledger records paid execution and applicable infrastructure costs. Ready-to-run examples cover common client languages, helping a developer move from an inspected playground response to a server-side request without translating the route contract from scratch.

For publishers, ApiClay can wrap sources that aren't already polished public APIs. It supports MySQL, PostgreSQL, MongoDB, Airtable, Google Sheets, Excel and CSV files, existing HTTP endpoints, and OpenAPI imports. A publisher defines routes, chooses which fields are exposed, maps a source response into a stable public shape, and sets access and pricing. Publisher Studio adds live testing, response mapping, encrypted drafts, and an optional AI assistant. Credentials are kept in server-side fields rather than placed in buyer clients. The source remains under the publisher's control, while a public route can expose only selected columns or fields. That lets a team offer a narrow data product without giving buyers direct access to the underlying database or provider account.

The website workflow goes beyond a basic HTML scraper. Live URL import can inspect background XHR and GraphQL traffic, create editable route drafts, and use static fetching, JavaScript rendering, or an automatic mode. Configured browser actions can navigate, fill inputs, click, wait, scroll, and capture structured responses. A separate extraction API returns page content as JSON, Markdown, text, or links. Monitoring can schedule checks against an API or webpage, track selected field changes, and send signed webhooks when something relevant changes. Publishers still need permission to retrieve and productize a source, and mappings must be retested when an upstream page or response changes. ApiClay manages the execution layer, but it can't make an unstable external layout maintenance-free.

ApiClay also exposes a hosted Model Context Protocol server for agents. An MCP-compatible client can search the marketplace, inspect an API schema and price, then call an approved route through one endpoint. Discovery tools are free, while execution uses the listing's billing rules. The same buyer key works for MCP and REST requests. This is useful when an agent needs current structured data but shouldn't receive a publisher's database password, proxy credential, or upstream API token. The client retains control over tool approval, so connecting the server doesn't grant the model unrestricted authority to spend credits. An agent can inspect the listing first, then the user or application can approve execution under the selected route's rules.

The platform fits several groups. Developers can add live data without building a separate proxy for every vendor. Startup teams can stand up an MVP backend from a sheet or database. Business teams can enrich uploaded lists and export the results with spending caps. Data owners can package a controlled subset of an internal source as a product. Agent builders can give a model a searchable tool catalog while preserving client-side approval over consequential calls. The broad surface is useful, though a team may need time to understand which source, browser, monitor, or marketplace workflow matches its task. It is strongest when a stable contract, controlled access, documentation, and metering matter as much as retrieving the data itself.

Access is freemium in practice. Browsing, search, documentation, and schema discovery are free, and individual listings may include a trial. Paid calls draw from a credit balance according to the provider's price, with possible extra charges for browser rendering, verified CAPTCHA handling, or proxy use. Publishers choose their own plans, per-call price, and limits. This flexibility is central to the marketplace model, but it also means there isn't one simple platform price to compare. Buyers should inspect the live Pricing tab for every route before moving a workload into production. A prototype that stays inside discovery can be inexpensive, while a browser-heavy integration may have a more complex cost profile.