Bright Data datasets in 2026

The first datasets on the marketplace, with their sizes

The marketplace opens with these datasets. Each order costs at least $250, which buys 100K records at $0.0025 a record, and the price per record falls with volume. Beside each, the SpecStream page for the same records as rows.

LinkedIn people profiles: 676.2M records. On SpecStream: the LinkedIn profiles dataset.

Amazon products: 758.5M records. Amazon Reviews: 43.1M records. On SpecStream: the Amazon reviews dataset.

LinkedIn company information: 30.4M records. Crunchbase companies information: 5.3M records. On SpecStream: the LinkedIn company dataset and the company data API.

Instagram profiles: 661.1M records. On SpecStream: the Instagram profiles dataset.

LinkedIn job listings information: 19.7M records. On SpecStream: the LinkedIn jobs dataset.

Google Maps full information: 269.4M records. On SpecStream: the Google Maps dataset.

Zillow properties listing information: 137.4M records. On SpecStream: the Zillow dataset.

YouTube videos posts: 4.1B records. On SpecStream: the YouTube videos dataset.

$250 is the floor, even for 556 records

Bright Data also sells datasets for single brands. The Transparent Labs dataset covers protein powders, pre-workouts and creatine, with prices, sale prices, flavors and reviews. It lists 556+ records, and the $250 minimum order makes that about 45¢ a record.

On SpecStream the request names the products and the facts, and each row is read from the store during your run. 556 rows with up to 5 facts cost $27.80 at 5¢ a row, and a row with up to 15 facts and its reviews costs 15¢. Below 5,000 rows at 5¢, a SpecStream run costs less than one $250 order.

Schedules and New Records: what an update costs on each side

A Bright Data dataset subscription refreshes on a schedule you pick: one time, biannually, quarterly, monthly, weekly or daily. You set the destination once under Control Panel, My Datasets, Delivery Settings, and the data is delivered automatically on each refresh. A snapshot stays available for 14 days after it is made.

The price per record falls with the schedule. From up to $0.0025 a record, refresh plans take 25% off biannual, 50% off quarterly and 80% off monthly, so a monthly refresh comes to $0.0005 a record. Smart data updates go further: you can buy only New Records or only Updated Records instead of the whole dataset again. Every Marketplace purchase still has a $250 minimum order.

Two costs sit in the purchase options. Records outside the time range you pick are scraped when the order runs and carry an additional fee. And when a record no longer matches your filters after that scrape, it is removed from the dataset and the scrape is still charged.

On SpecStream the schedule belongs to the request. A saved request reruns hourly to yearly inside the spending limits you set. Each row comes back labeled changed, unchanged or no longer matching: a changed or new row costs half its rate, an unchanged row costs 1¢, and a row a page contradicts is not charged. New matches go to Slack, email, a sheet, a CRM or a signed webhook.

For the 556 Transparent Labs products: the first run costs $27.80 at 5¢ a row. A monthly rerun where 30 rows changed and 526 did not costs 30 rows at 2.5¢ plus 526 rows at 1¢, $6.01 in all. The nearest Bright Data order is $250 at the start, whatever the refresh plan.

For public records there is a third way: many of the ready lists carry a monthly feed that delivers the rows the source adds each week, at the list's own two rates on the rows added in a month, with a $19 a month minimum.

Pre-built scrapers: $1.5 per 1,000 records, for URLs you already have

A Bright Data scraper takes a URL and returns one structured record for it, at $1.5 per 1,000 records pay as you go or $499 a month for 384,000. Finding the URLs stays with your code. A SpecStream request finds the pages itself, and the searches and pages behind each row are not billed.

When you have the URLs and want the page text, the Extract API reads 1,000 pages for $0.80 as Markdown, up to 20 a call, and a page with no text is not charged.

Ready lists of public records from $5

On Nov 12, 2026, SpecStream sells 100,666 ready lists from 46 official sources, 91,961,979 rows in all: nonprofits, UK and France companies, store locations and more. A list costs 5¢ a row for the first 2,000 rows and 1¢ a row after, with a $5 minimum, paid by card with no account. Every list is on the data lists page.

Scored on the official benchmarks

The SpecStream Search API scores 97.9 F-score on all 1,000 SimpleQA Verified questions, 978 of them correct, and 64.0 item F1 and 43.4 row F1 on all 200 WideSearch tasks. The Extract API scores 0.908 F1 on WCXB. Every score and its method is on the benchmarks page.

SpecStream

The live web data engine for teams and agents. One request in plain words returns every match on the live web as structured rows at 5¢, 15¢ or $1 a delivered row, under a maximum you approve. Ready lists of public records cost 5¢ a row for the first 2,000 rows and 1¢ after, and a new account starts with $10 of credit, no card.

Oxylabs

A Web Scraper API billed per result, from $0.50 per 1,000 results, with $5 of welcome credit toward any product and Enterprise plans starting at $2,500 a month. Each request names one URL or query. SpecStream returns the finished rows, with no URL list to build first.

Apify

A store of scrapers called Actors, each priced by its author. Its Instagram Scraper charges $2.30 for 1,000 results, and the Starter plan costs $19 a month. SpecStream has one rate card for every request.

Firecrawl

Credits on monthly plans: 1 credit is 1 page on a basic scrape, the JSON format adds 4 credits per page, and Hobby gives 5,000 credits a month. SpecStream prices the structured row, not the page read to fill it.