Frequently Asked Questions

Search API is a no-code platform that turns your data into a production-ready REST API backed by Elasticsearch. You define models and endpoints in the dashboard, and Search API generates the API, its authentication and its Swagger documentation for you.

No. Normally querying Elasticsearch means writing JSON Query DSL by hand; Search API generates those queries for you. Searching, filtering, sorting and pagination are exposed as simple URL query parameters on every endpoint.

No. Every API uses the Elasticsearch cluster hosted by Search API by default. If you already run Elasticsearch, you can connect your own self-managed or Elastic Cloud cluster with a username and password or an API key instead.

Search API imports CSV and Excel (XLSX) files, as well as ZIP and TAR archives that contain one of them, up to 100 MB. It can create a new model, mapping and endpoint from the file automatically, or update the records of an existing model.

Yes. SQL Import reads a mysqldump file of up to 100 MB and turns every table into a model with its own endpoints and Elasticsearch index. Foreign keys become relations between models, and many-to-many join tables become nested fields.

Yes. Any model can be exported as CSV, JSON or SQL at any time from the Data UI. The SQL export contains CREATE TABLE and INSERT statements that load straight into a MySQL-compatible database, so your data is never locked in.

Every endpoint lives at https://{subdomain}.searchapi.net/{api-path}/{version}/{endpoint-path} and requires an X-API-KEY header. List endpoints support full-text search with q, exact and range filters, sorting and field selection with includes and excludes.

Access is granted per API key and per endpoint, so a key can be allowed to read with GET but not to write with PUT or DELETE. Requests without a valid key are rejected, and you can grant a key access to a single API version or to all versions at once.

Yes. Each API version is an independent snapshot with its own URL segment such as v1 or v2, and a deployed version is locked. When you deploy a new version, records are copied from the previous version automatically while existing clients keep using the old one.

Yes. Every deployed API version gets an OpenAPI 3.0 (Swagger) document that you can view, download or share with your team. Endpoints can be tested directly in the browser from that document.

Yes. Search API runs an MCP (Model Context Protocol) server, so assistants such as Claude Code, Cursor and Gemini CLI can create models, APIs, endpoints and API keys and deploy versions in plain language. It authenticates with a personal access token from your account settings.

Yes. Every new workspace starts on the Free plan, and paid plans with higher API, model, data and daily request limits are available monthly or yearly. Payments are processed by Stripe, and Enterprise plans with custom limits are available on request.

You receive email notifications when your workspace reaches 90%, 100% and 110% of its daily request limit. At 110%, API calls are blocked until the daily counter resets, or until you upgrade to a plan with a higher limit.

Yes. Besides the self-serve platform, the Search API team offers Elasticsearch consulting, cluster implementation, search relevancy tuning and 24/7 support, based on nearly 12 years of Elasticsearch experience.