Amazon Rufus
Extract structured product recommendations from Amazon Rufus - ASINs, ratings, editorial tags, and follow-up questions - as clean JSON.
Overview
Use this endpoint to monitor which products Amazon’s AI recommends for any category, track competitor product placements in Rufus answers, and research what editorial tags (like “Best overall” or “Best value”) Amazon assigns to products in your space.Unique features
- ASINs: Amazon Standard Identification Numbers for every recommended product - enabling direct product tracking
- Editorial tags: Rufus’s curated labels like “Best overall”, “Best value”, or “Most reviewed”
- Related questions: Follow-up questions Rufus suggests - useful for understanding shopping intent
- Multi-marketplace: Automatic routing to the correct Amazon marketplace based on country code
Supported marketplaces
Example request
Response schema
Includes common response fields plus:Common questions
Why does Amazon Rufus cost 4 credits instead of 3?
Rufus requires authenticated Amazon sessions with anti-bot token handling and real-time SSE stream parsing - more infrastructure than a standard web scrape. The extra credit reflects this complexity.Are all product fields always populated?
No. Fields likerating, reviews, image_url, category, url, and footer are nullable - Rufus doesn’t always return all fields for every product. Always check for null before processing.
How do I track if my product appears in Rufus recommendations?
Query your product category and checkproducts[].asin for your ASIN. Run the same prompt repeatedly across different countries to monitor regional placement and track changes over time.
Which marketplaces support all features?
All supported marketplaces return ASINs, titles, and follow-up questions. Rating, review count, and editorial tags depend on the marketplace’s Rufus implementation and may vary by region.Authorizations
Your ScrapeLLM API key passed in the request header.
Query Parameters
The shopping query to send to Amazon Rufus. Free-form natural language: direct questions, product comparisons, or shopping intent. Maximum 4,000 characters.
1 - 4000ISO 3166-1 alpha-2 country code for the Amazon marketplace.
Maps to the appropriate domain: US → amazon.com, GB → amazon.co.uk,
DE → amazon.de, ES → amazon.es, etc. Defaults to US.
2Skip the response cache and always fetch a fresh result.
Maximum seconds to wait for Rufus to respond. Must be between 10 and 600.
10 <= x <= 600Response
Scrape completed successfully.
Always "amazon_rufus".
"amazon_rufus"
"done" on success.
"done"
Unique identifier for this scrape job.
"3fa85f64-5717-4562-b3fc-2c963f66afa6"
The shopping query that was submitted.
"best wireless headphones under $100"
The Amazon marketplace country targeted.
"US"
Rufus's curated product list for the query. Typically 3–6 items. Empty array if Rufus found no recommendations.
Follow-up shopping questions generated by Rufus. Useful for query expansion, SEO keyword research, or chained calls.
Credits consumed by this request. Always 3.
3
Total end-to-end request duration in milliseconds.
8341.2
true if this response was served from cache.
false
Loading-state messages localised to the marketplace.
Hosted .md summary URL for this response. Expires after 24 hours.
"https://scrapellm.com/md/3fa85f64-5717-4562-b3fc-2c963f66afa6.md"
API version that produced this response.
"1.0.0"
ISO 8601 UTC timestamp when the gateway processed this response.
"2026-05-27T12:00:00+00:00"
ISO 8601 UTC timestamp when this response was originally cached.
Only present when cached is true.
"2026-05-27T12:00:00+00:00"