CarBuzz Scraper

Extract structured vehicle-review data from public CarBuzz make/model/year pages. Supply one or more CarBuzz vehicle URLs and receive normalized records with pricing, BuzzScore ratings, rating breakdowns, engines, drivetrains, transmissions, horsepower, torque, fuel economy, pros, cons, base specifications, and published trim cards.

This Actor is an independent community tool. It is not affiliated with, endorsed by, or operated by CarBuzz or Valnet. Review text, ratings, images, and vehicle specifications remain attributable to their source. Follow applicable terms and use the data responsibly.

Input

startUrls accepts public HTTPS URLs under carbuzz.com/cars/, for example:

{
  "startUrls": [
    "https://carbuzz.com/cars/toyota/camry/2025/"
  ],
  "maxPages": 1,
  "maxTrimsPerVehicle": 20
}

The Actor normalizes and deduplicates URLs before fetching. maxPages limits unique vehicle pages. maxTrimsPerVehicle caps trim cards attached to a result. maxConcurrency is intentionally limited to three, while requestDelayMillis enforces pacing. maxRetries, maxPageMbytes, and maxRunMillis provide bounded failure and resource controls. Apify Proxy is optional; direct public access is used by default.

Output

Each dataset row represents one canonical vehicle-year review and includes:

• stable hashed recordId, vehicle name, model year, make, model, and body type;
• source summary and primary image URL;
• engine, drivetrain, transmission, horsepower, torque, and fuel economy when published;
• numeric starting price and currency;
• numeric overall BuzzScore and named rating components;
• ordered pros and cons from CarBuzz structured review data;
• baseSpecifications with published base-trim labels and values;
• trims, where each trim contains its name and the specifications displayed on its card;
• review author and publication date when available;
• canonical source URL and scrape timestamp.

Unavailable optional keys are omitted. Numeric price and rating fields remain JSON numbers; specification strings retain source units and qualifiers.

Extraction and data quality

CarBuzz exposes rich schema.org structured review data in its public HTML. The Actor uses that structured data for vehicle identity, offer price, rating, rating breakdown, author, publication date, pros, and cons. It supplements this with public base-trim and trim-card specification pairs from the rendered HTML.

Every page must return valid HTML and a real vehicle Review object. The Actor rejects challenge pages, oversized responses, missing required identity fields, and duplicate canonical URLs. It uses realistic consistent headers, one anonymous public request per page, responsible pacing, bounded exponential retries, and a deadline below five minutes. It never creates mock vehicles, placeholder values, or hardcoded success records.

Limitations

Specifications and ratings reflect CarBuzz editorial content at scrape time and may differ by market, drivetrain, trim, options, or later correction. A page may not publish every optional field. Trim cards shown in an article can be a curated subset rather than the entire manufacturer configuration catalog. Pricing generally excludes taxes, registration, destination fees, and optional equipment unless the source says otherwise.

Use manufacturer documentation for purchase-critical configuration details and safety recalls. CarBuzz reviews are editorial opinions, not mechanical inspections or guarantees.