Skip to content

ChocoData-com/amazon-scraper-api-sdk-python

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

amazonscraperapi-sdk

PyPI version PyPI downloads license

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:

Metric Value
Median latency (product, US) ~2.6 s
P95 latency ~6 s
P99 latency ~10.5 s
Price / 1,000 Amazon products $0.50 flat
Concurrent threads (entry paid plan) 50
Marketplaces supported 20+
Billing unit per successful (2xx) response

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"])
# "Apple AirPods Pro (2nd Generation)..."
print(product["price"]["current"])
# 199.00
print(product["rating"]["average"], product["rating"]["count"])
# 4.7 58214

Example output (trimmed)

{
    "asin": "B09HN3Q81F",
    "title": "Apple AirPods Pro (2nd Generation)...",
    "brand": "Apple",
    "price": {"current": 199.00, "currency": "USD", "was": 249.00, "savings_pct": 20},
    "rating": {"average": 4.7, "count": 58214, "distribution": {"5": 0.81, "4": 0.12}},
    "availability": "In Stock",
    "buybox": {"seller": "Amazon.com", "ships_from": "Amazon.com", "prime": True},
    "images": ["https://m.media-amazon.com/images/I/...jpg"],
    "bullets": ["Active Noise Cancellation...", "Adaptive Audio..."],
    "variants": [{"asin": "B0BDHB9Y8H", "name": "USB-C", "price": 249.00}],
    "categories": ["Electronics", "Headphones", "Earbud Headphones"],
    "_meta": {"tier": "direct", "duration_ms": 2634, "marketplace": "amazon.com"},
}

Keyword search

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

for r in results["results"]:
    print(r["position"], r["asin"], r["title"], r["price"].get("current"))

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:

Pain point What we handle
Amazon CAPTCHAs / robot pages Auto-detected, retried through a heavier proxy tier (datacenter, residential, premium)
Brittle CSS selectors Extractors update as Amazon changes layouts. Your code stays the same.
20+ marketplaces amazon.de, .co.uk, .co.jp, .com.br, and more. Each with marketplace-specific parsing quirks.
Country-matched residential IPs amazon.de auto-routes through German IPs. Pass country="DE" to override.
Rotating proxies + anti-fingerprinting TLS fingerprints, headers, cookies handled server-side.
Rate-limit retries Transparent exponential backoff.
Structured JSON output Title, price, rating, reviews, variants, seller, images. Parsed, typed.
Batch/async jobs 1,000 ASINs submitted, webhook-delivered on completion.

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

All failures follow a stable shape so you can match on code:

from amazonscraperapi import AmazonScraperAPI, AsaError
import time

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

try:
    product = asa.product(query="INVALID_ASIN", domain="com")
except AsaError as e:
    if e.code == "INSUFFICIENT_CREDITS":
        pass  # top up
    elif e.code == "RATE_LIMITED":
        time.sleep(e.retry_after)
    elif e.code in ("target_unreachable", "amazon-robot-or-human"):
        # non-2xx, you were not charged. safe to retry.
        pass
    else:
        raise
HTTP code When you see it Recommended action
400 INVALID_PARAMS Missing query, unsupported domain, bad sort_by Fix request; don't retry
401 INVALID_API_KEY Missing, malformed, revoked Verify ASA_API_KEY; rotate if leaked
402 INSUFFICIENT_CREDITS Balance empty Top up; renews each billing cycle
429 RATE_LIMITED Over 120 req/60s per user Honor Retry-After; retry
429 CONCURRENCY_LIMIT Over plan's parallel cap Reduce parallelism or upgrade. X-Concurrency-* headers guide backoff
502 target_unreachable Amazon down / all proxy tiers blocked Retry after 30s (already retried through 3 tiers on our side)
502 amazon-robot-or-human Amazon challenge not resolvable Retry. Often transient. Not charged
502 extraction_failed Layout we can't parse Report with X-Request-Id. Not charged
503 SERVICE_OVERLOADED Global circuit breaker Honor Retry-After: 60. Rare
500 INTERNAL_ERROR Our bug Report with X-Request-Id

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

License

MIT

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages