API connectors

Vendors

ActiveCampaign, Affinity, Ahrefs, Alpha Vantage, Apollo.io, Appfigures, Binance, Bing Webmaster Tools, Blueshift, Braze, Census, Cloudflare Radar, CoinAPI, CoinGecko, CoinMarketCap, Commodities-API, Copper, CryptoCompare, DataForSEO, Docker Hub, Drift, Electricity Maps, EODHD, Financial Datasets, Financial Modeling Prep, Finnhub, Fixer, Folk, Frankfurter, FRED, Freshdesk, Freshsales, GitHub, Gorgias, GTmetrix, Hightouch, Insightly, Iterable, Keepa, Klaviyo, Kustomer, Less Annoying CRM, Libraries.io, Lytics, Mailchimp, Majestic, Marketstack, Metals-API, Moz, npm, Nutshell, OneSignal, Open Exchange Rates, Open-Meteo, OpenWeatherMap, PageSpeed Insights, Polygon.io, PyPI, Salesflare, SE Ranking, Semrush, SendGrid, SerpApi, Serpstat, Similarweb, SpyFu, Streak, Tiingo, Tomorrow.io, Trading Economics, Treasure Data, Twelve Data, U.S. Bureau of Labor Statistics, U.S. Energy Information Administration, ValueSERP, Visual Crossing, WeatherAPI.com, World Bank Open Data, Yandex Webmaster and Zendesk.

They appear under Apps in the Add Data Source dialog with a REST label. They are available on every plan. A few need no key (Binance, Docker Hub, Frankfurter, npm, Open-Meteo, PageSpeed Insights, PyPI, World Bank Open Data); for GitHub a token is optional.

Connect

1. Open Settings > Integrations, click Add Data Source and search for the vendor.
2. Enter the API key (or the credentials the vendor needs, such as a login and password). Vendors whose address depends on your account also ask for it, for example your Zendesk subdomain. Name the source and click Continue.
3. Under Choose what to sync, fill in the lists the vendor works with: stock symbols, coins, domains, keywords, projects. Some lists search the vendor as you type; others take free text with Add. Then set the options the vendor offers, such as markets, languages or modules.
4. Click Create Data Source. The first sync runs straight away and a summary shows what arrived.

Keys are encrypted and used only to pull the data you selected. The connector only reads.

Vendors that charge per request

For vendors that bill each request, such as DataForSEO, the dialog shows Estimated per sync: the number of requests and, where the vendor publishes prices, the approximate cost. The vendor bills you, not clariBI. These vendors also have a Refresh cadence, weekly by default. Some requests are queued at the vendor's cheaper standard tier; their results arrive a few minutes after the sync.

More endpoints from the vendor's catalog

On the trial, Starter, Professional and Enterprise, Additional endpoints lists every read endpoint the vendor documents (79 vendors have a catalog, 4,352 endpoints in all). A source can use up to 30.

1. Click Add an endpoint, search by name, section or path, and click Configure.
2. Set Name on the dashboard, History (Keep every sync or Latest sync only) and, for dated data, Look-back (days).
3. Fill the endpoint's fields. Most can take their values from the lists above (for example "each of your domains"). Rarely used fields sit under More options.
4. For live endpoints, Preview (1 request) fetches a sample first. The editor shows how many requests the endpoint will make per sync.

Dates are never frozen at the day you set an endpoint up. Every date field takes its value from the source's period: the start or end of the look-back window, each day in it, or a number of days after the sync. Each endpoint syncs into its own dataset.

Editing a setup

In Settings > Integrations, open the source's Manage menu and choose Edit Configuration. The Edit Setup section changes the name, description, lists, options, sync cadence, credentials and additional endpoints without deleting data already synced. Credentials left empty keep the stored ones. Change history lists earlier edits with who made them. Save, or Save & sync now to fetch with the new setup at once. The last sync's summary is shown at the bottom.

How syncs work

• A source syncs when you click Sync now or Sync Data, and when something uses it and its data is older than its cadence (24 hours by default).
• Dated data is fetched incrementally: from three days before the newest day already stored, so late corrections are picked up.
• Synced rows go through the same processing as other sources, so the source gets a Data Insights dashboard and can be used in questions, dashboards and reports.

Custom API and web scraping

For a tool without a connector, the Custom group has two options.

Custom API connects one REST endpoint that returns JSON:

• API URL and Endpoint path
• Authentication Type: No Authentication, Bearer Token, API Key (with the key's parameter name) or Basic Auth
• Method: GET or POST
• Query parameters, one name=value per line
• Records path: where the list of records sits in the response, for example data
• Pagination: none, a page number parameter, or offset and limit parameters, with the page size

Web Scraping reads data from a web page, either as static HTML or with JavaScript rendering, using the selectors you give it.

If the vendor runs an MCP server, a custom MCP connection (Starter and up) may suit better; see MCP integrations.