amazonscraperapi-sdk

Official Python SDK for Amazon Scraper API. Pay only for successful (2xx) responses — non-2xx never cost you anything. Pricing starts at $0.90 per 1,000 requests on pay-as-you-go, drops to a flat $0.50 per 1,000 requests on Custom plans. 1,000 free requests on signup, no card. Drop into any Python project to fetch structured Amazon product data, run keyword searches, or queue async batches with webhook callbacks.

Benchmark (live production, 2026-04)

Measured on our own infrastructure against a 30-query mixed international set:

Install

pip install amazonscraperapi-sdk

Requires Python >= 3.9. Built on httpx.

Quick start - single product

from amazonscraperapi import AmazonScraperAPI

asa = AmazonScraperAPI(api_key="asa_live_...")

product = asa.product(query="B09HN3Q81F", domain="com")

print(product["title"])
# "LEGEND COOKWARE 5-Ply Stainless Steel Cookware Set, 14-Piece..."
print(product["price"], product["currency"])
# 549.99 USD
print(product["rating"], product["reviews_count"])
# 4.5 865

Example output (trimmed, ~55 fields in total)

{
    "asin": "B09HN3Q81F",
    "title": "LEGEND COOKWARE 5-Ply Stainless Steel Cookware Set, 14-Piece...",
    "brand": "LEGEND COOKWARE",
    "price": 549.99,
    "price_strikethrough": None,
    "currency": "USD",
    "rating": 4.5,
    "reviews_count": 865,
    "stock": "In Stock",
    "is_prime": True,
    "images": ["https://m.media-amazon.com/images/I/71Q3a8JH5NL._AC_SY355_.jpg", "..."],
    "bullet_points": "5-PLY STAINLESS STEEL COOKWARE HEATS EVENLY - Five layers of steel...",
    "sales_rank": [
        {"ladder": [{"name": "Kitchen & Dining", "url": None}], "rank": 36177},
        {"ladder": [{"name": "Kitchen Cookware Sets", "url": None}], "rank": 174},
    ],
}

Keyword search

results = asa.search(
    query="wireless headphones",
    domain="co.uk",
    sort_by="avg_customer_review",
    pages=1,
)

for p in results["products"]:
    # Sponsored cards have sponsored_position instead of organic_position.
    print(p["organic_position"], p["asin"], p["title"], p["price"], p["rating"])

Async batch (up to 1,000 ASINs with webhook callback)

batch = asa.create_batch(
    endpoint="amazon.product",
    items=[
        {"query": "B09HN3Q81F", "domain": "com"},
        {"query": "B000ALVUM6", "domain": "de", "language": "de_DE"},
        # ... up to 1,000
    ],
    webhook_url="https://your.server/webhooks/asa",
)

print("batch id:", batch["id"])
# SAVE THIS. Webhook signing secret is returned only once:
print("webhook secret:", batch["webhook_signature_secret"])

# Alternative: poll
status = asa.get_batch(batch["id"])
print(f"{status['processed_count']}/{status['total_count']}")

Verifying webhook signatures

from amazonscraperapi import verify_webhook_signature
from fastapi import FastAPI, Request, HTTPException
import json, os

app = FastAPI()

@app.post("/webhooks/asa")
async def asa_webhook(request: Request):
    raw = await request.body()
    signature = request.headers.get("X-ASA-Signature")
    if not verify_webhook_signature(signature, raw, secret=os.environ["WEBHOOK_SECRET"]):
        raise HTTPException(401, "invalid signature")
    payload = json.loads(raw)
    # process payload["results"]

What the API solves for you

Building a production-grade Amazon scraper in-house is a 2-4 week engineering project plus permanent maintenance. This SDK wraps Amazon Scraper API, which has already solved:

Time saved: a greenfield Python Amazon scraper built to this spec takes roughly 80 engineer-hours (including anti-bot handling and marketplace variants). This SDK is 10 minutes.

Error handling

Every failed call raises AmazonScraperAPIError with status_code, code (the API's error code, or None) and body (the parsed JSON body). The SDK does not retry on its own, so match on e.code and decide:

import time
from amazonscraperapi import AmazonScraperAPI, AmazonScraperAPIError

asa = AmazonScraperAPI(api_key="asa_live_...")

try:
    product = asa.product(query="B09HN3Q81F", domain="com")
except AmazonScraperAPIError as e:
    if e.code == "INSUFFICIENT_CREDITS":
        pass  # top up
    elif e.code == "RATE_LIMITED":
        time.sleep(e.body["error"].get("retry_after_ms", 1000) / 1000)
    elif e.code == "target_unreachable":
        time.sleep(60)  # not charged; retry once
    else:
        raise

The default request timeout is 120s (the API's own ceiling); override it with AmazonScraperAPI(api_key=..., timeout=...).

Flat-credit promise: non-2xx responses are free. A basic request costs 5 credits (billing unit; end-customer price is $0.90 per 1,000 basic requests PAYG). Future JS-rendered calls will cost 15 credits. Every response has X-Request-Id for traceability — quote it in any support ticket.

Get an API key

app.amazonscraperapi.com. 1,000 free requests on signup, no credit card required.

Links

• Website: https://www.amazonscraperapi.com/
• Docs: https://amazonscraperapi.com/docs
• Pricing: https://amazonscraperapi.com/pricing
• Node SDK: amazon-scraper-api-sdk · Go SDK: github.com/ChocoData-com/amazon-scraper-api-sdk-go · CLI: amazon-scraper-api-cli · MCP server: amazon-scraper-api-mcp

License

MIT