Getting Started

Create your first AI pipeline in under 5 minutes. This guide walks you through account setup, writing your first pipeline definition, running it, and viewing the results.

Prerequisites

Before you start, you'll need:

• A stepIQ account (free tier works)
• Basic understanding of YAML or JSON syntax
• An idea for an AI workflow (or just use our example)

Step 1

Create an account

Sign up at app.stepiq.io with your email address. You'll get 100 free credits to start — enough to run several pipeline tests.

Step 2

Write your first pipeline

Create a new pipeline from the dashboard, or write a YAML definition. Here's a minimal example:

name: "summarize-article"
version: 1

input:
  schema:
    article_url:
      type: string

steps:
  - id: summarize
    model: claude-sonnet-4
    prompt: |
      Summarize this article in 3 bullets:
      {{input.article_url}}
    output_format: markdown

Step 3

Run the pipeline

Click "Run" from the dashboard or trigger via API. Provide the required inputs and watch the execution in real-time with SSE streaming.

Trigger methods

Inbound webhook quickstart

Create an API key in Settings → API Keys, then trigger your pipeline from any external HTTP client:

Plan note: webhooks are available on starter, pro, and enterprise plans, and not available on free.

curl -X POST "https://api.stepiq.sh/v1/api/webhooks/<pipeline-id>" \
  -H "Content-Type: application/json" \
  -H "X-API-Key: sk_live_xxxxxxxx" \
  -d '{"input_data":{"article_url":"https://example.com/blog-post"}}'

The API returns 202 Accepted with a run_id you can track via GET /api/runs/:id.

Step 4

View results

Each run shows a detailed step-by-step execution log. For every step you can see:

• The interpolated prompt sent to each model
• Raw and parsed output (JSON or markdown)
• Token counts (input + output) and cost in credits
• Duration, retry count, and status per step

Next steps