Skip to main content
A workflow is a saved pipeline of blocks that finds, researches and delivers records. Build it once, then run it on a file, a webhook or a schedule. Each run is a Job.

Core concepts

Execution flow

  1. List workflows to find the workflow_id you want to run.
  2. Run the workflow with input rows and receive a Job.
  3. Poll the Job until status is succeeded, failed or canceled.
  4. Download the output files listed in the Job’s result.
To build workflows without code, use the Workflow Editor in the Sixtyfour web app.

Run a workflow

Replace YOUR_API_KEY with your API key and WORKFLOW_ID with a workflow_id from the list. cURL uses jq; Python uses requests; JavaScript uses Node.js 18 or later.

List workflows

HTTP 200

Start a run

Send the rows in input, one object per row. input applies to workflows whose first block is Incoming Webhooks. For a workflow that reads a file, see Run on an uploaded file.
The response is a Job:
HTTP 202
To retry a start safely, send an Idempotency-Key header. A retry with the same key and body returns the original Job.

Poll the Job

A successful workflow Job lists its output files in result, one per block and format:
HTTP 200 (excerpt)

Download the output

Fetch each download_url with your API key. In a linear workflow, the highest block_number is the final output; a branching workflow can have several output files. For complete download code in each language, see Download Workflow output.
cURL
Download links expire. Read the Job again to get fresh ones.

Run on an uploaded file

A workflow whose first block is Upload Data (read_csv) reads its rows from a file. Upload the file with POST /v1/files, then pass the returned file_id as source_overrides.resource_handle_id:
cURL
Files can be CSV, JSON, JSONL or NDJSON. See Run a workflow batch for Python and JavaScript.
Researching a list of people or companies? Use Bulk Intelligence: one call with a file and fields, no workflow needed.

Output formats

CSV is the default. To also save NDJSON, send result_formats when starting the run. NDJSON keeps numbers, booleans and nested objects typed, where CSV writes every value as text.
cURL
The finished Job’s result then lists a file for each format. Select the one you need by content_type.

Errors

For error responses, see Handling Errors.

Pricing

See Credit Costs for more information.
Integrations built on /workflows/run and the other original routes keep working. Legacy maps each one to its /v1 replacement.