Databar

In SpecStream, open Agents & API, choose API keys, then Create API key, and name the key Databar.

The Search API needs Authorize paid work. Deep Search also needs Read saved requests, Create and edit requests, Prepare work and follow-ups, and Read results and their sources. Under Authorize paid work, give the key a Total spending limit. Before the key can start a run, set your account's own spending limit in Settings, under Usage and billing.

The key is shown once, so copy it before you close the dialog.

In Databar, open Integrations from your workspace home and click Connect a custom API. Fill it in as below: for authentication, send the header Authorization with Bearer and your key, and add one request parameter, q, for the question. Click Save.

Method: GET
URL: https://api.specstream.ai/v1/web-search
Header: Authorization = Bearer YOUR_SPECSTREAM_KEY
Parameter: q

In a table, add a column that holds the question for each row, such as Who is the chief executive of Datadog?. In the Enrich sidebar, choose Add or use your own REST API, pick the search connector, and map the question column to q. Click Add Columns, then Run.

data.answer.value is the answer, data.answer.page is the page it was read from, and data.answer.seen_at is when it was read. When no result prints an answer, data.answer is empty and data.sources lists the results read.

Each search is $5 per 1,000 searches of up to 10 results, with no quote and no approval.

Deep Search finds every match for a request. Connect a second custom API with the method POST, the URL https://api.specstream.ai/v1/capacity/requests, and the same Authorization header. Add a parameter Content-Type with the value application/json, which takes the place of Databar's default.

Under Body parameters, click Use a custom template and paste the template below. Databar puts each row's value in place of each name between dollar signs. requested_units is how many results to find, and confirm_max_cents is the most the request may cost, in cents: here 25 results for at most $5. Click Save.

{
  "q": "$request$",
  "requested_units": 25,
  "confirm_max_cents": 500,
  "idempotency_key": "$key$"
}

In a table, add a column with the request, such as Cybersecurity software companies in Germany with 50 to 200 employees, with their website and headquarters city, and a column with a key of 8 to 128 characters, new for each request. Add the Deep Search connector, map the columns to request and key, and run it.

SpecStream answers at once with data.state set to reading. It then reads the request and prices it, and starts the run only when the price is at or below confirm_max_cents. Above it, nothing runs.

To see the price before anything runs, set confirm_max_cents to 1 in the template. The state is then above_maximum, and the next step shows the price. To approve that price, put it in the template and start the row again with a new key.

Connect a third custom API with the method GET, the same Authorization header, and the URL below, with the request's key in place of YOUR_REQUEST_KEY. Reading it costs nothing.

data.state says where the request stands, data.read_back is how SpecStream read it, and data.maximum_cents is the price. Run it again until data.done is true. The rows are then in data.flat_rows, one item per result, up to 100 a call. For more, add the parameter cursor with the value of data.next_cursor.

The same rows open in your SpecStream workspace at list_link, where Export downloads them as a CSV.

https://api.specstream.ai/v1/capacity/requests/YOUR_REQUEST_KEY/rows