SupaCrawlX Python SDK: Complete Quickstart Guide

The fastest way to get from zero to data is the official SDK. It handles auth, retries, and typing so you can focus on what you actually want to build.

Install

pip install supacrawlx

Authenticate

Grab a key from your dashboard and pass it once when you create the client:

import os
from supacrawlx import Client

client = Client(api_key=os.environ["SUPACRAWLX_API_KEY"])

Transcripts

# Auto-detects the platform from the URL
yt = client.youtube.transcript(url="https://youtu.be/dQw4w9WgXcQ")
tiktok = client.transcript(url="https://www.tiktok.com/@user/video/730...")

Metadata, search, and channels

meta = client.metadata(url="https://youtu.be/dQw4w9WgXcQ")
results = client.youtube.search(query="vector databases", limit=25)
channel = client.youtube.channel(handle="@exampledev")

Web extraction

page = client.web.scrape(url="https://example.com")    # clean Markdown
sitemap = client.web.map(url="https://example.com")    # all URLs
job = client.web.crawl(url="https://example.com", limit=200)

Video analysis

summary = client.video.analyze(
    url="https://youtu.be/your_video",
    prompt="Summarize the key takeaways as bullet points.",
)

Handle errors cleanly

Every failure raises a typed APIError with a status code and machine-readable code:

from supacrawlx import APIError

try:
    client.youtube.transcript(url=bad_url)
except APIError as e:
    print(e.status_code, e.code, e.message)

Read credit usage straight from the response headers (X-Credits-Used, X-Credits-Remaining) to build cost dashboards without a second API call.

Production checklist

Store your key in an environment variable or secret manager — never in source.
Set a sensible timeout and retry idempotent reads on 429/5xx.
Cache transcripts and metadata; they rarely change.