1 to 2 credits
Extract one page
One credit for a page, two through the datacenter proxy. That is about $1.20 or $2.39 per 1,000 pages on Startup.
Diffbot Pricing, Every Plan and the Credit Cost per Page
6 min read
Diffbot pricing starts free at 10,000 credits a month, then jumps to $299 a month for 250,000 credits and $899 for 1,000,000. One extracted page is one credit, so Startup works out to about $1.20 per 1,000 pages. The catch is the plan floor and the developer time around it: Crawl, which finds product URLs on its own, only arrives at the $899 tier.
figures taken from diffbot.com/pricing; prices in USD before tax
Robot console
No account needed Running · s Complete · rows
The plans
Diffbot plans and what each one includes
Every plan is a monthly pool of credits shared by all four APIs. The differences that matter are the rate limit
and whether Crawl is switched on.
| Plan | Price a month | Credits | Rate limit | Crawl | Pages if all credits go to extraction |
|---|---|---|---|---|---|
| Free | $0 | 10,000 | 5 a minute | No | About 10,000 |
| Startup | $299 | 250,000 | 5 a second | No | About 250,000 |
| Plus | $899 | 1,000,000 | 25 a second | Yes, 25 active | About 1,000,000 |
| Enterprise | Custom | Custom | 25+ a second | Yes, 100+ active | Custom |
Startup includes 250,000 credits with overage at $0.001 each. Plus adds 3 user seats and overage at $0.0009. Free needs no credit card. Billing is monthly, and Enterprise is a sales conversation covering custom volume, rate limits and support.
What a credit buys
Why a page is cheap and a company record is not
Diffbot prices actions, not pages, and the Knowledge Graph actions cost 25 to 100 times a page.
25 credits
One Knowledge Graph entity
Exporting or enhancing a company or person record costs 25 credits, or 100 for a facet record or a refresh. Startup covers about 10,000 entities.
0 credits
Crawl to find pages
Spidering a site for links costs nothing, but it needs Plus. Without it you supply every URL yourself.
Worked bills
Three monthly Diffbot bills, with the flat-plan alternative
Thirty days at the published rates. Your own pages may need a retry or a proxy, so test on the Free plan before you
size a tier.
Bill A · known URLs
300 product URLs, checked daily
9,000 credits a month, inside Free at $0. It works if a developer writes the calls and stores the JSON. With nobody to build that, WebRobot Launch runs the same list for $79 and fills your sheet.
Bill B · discovered catalog
20,000 competitor products, refreshed weekly
About 80,000 credits, which Startup covers, but finding the product URLs needs Crawl and therefore Plus at $899. On WebRobot the robot follows the category and pagination itself: 20,000 records a week fits Scale at $249.
Bill C · company data
8,000 companies enriched
At 25 credits an entity that is 200,000 credits, inside Startup at $299. We do not sell a company database, so for firmographics Diffbot is the right buy. Scanned PDFs and invoices are a different job, and extracting fields from scanned documents is built for it.
Compare
Diffbot pricing next to a flat record plan
For page extraction the question is not the price of a credit but what you pay before the first row arrives.
| Monthly volume | Diffbot | WebRobot |
|---|---|---|
| 5,000 pages, known URLs | Free, $0, you build the scheduler | Launch, $79 or $39 billed yearly |
| 10,000 pages, known URLs | Free, near the 10,000 credit cap | Launch, $79 or $39 billed yearly |
| 50,000 pages | Startup, $299 | Scale, $249 or $124 billed yearly |
| Any volume, URLs discovered by crawling | Plus, $899 | Launch or Scale, crawling included |
| Company and people records | Startup or Plus, 25 credits each | Not offered |
WebRobot counts a record once however many pages the robot read to fill it, and the plans include scheduling, logins and the REST API. The side-by-side, including where Diffbot wins, is on the Diffbot alternative page. For the wider market see the best web scraping tools roundup, and for typical budgets, how much web scraping costs.
Who should buy what
Which plan fits which team
Buy Diffbot Startup or Plus
A product team with developers that wants JSON in Diffbot's schema inside its own application, or needs Knowledge Graph company data.
Start on Diffbot Free
An engineer testing whether automatic extraction handles your page types, up to 10,000 credits and 5 requests a minute.
Skip the credits
The job is a fixed set of fields from competitor or supplier sites, on a schedule, in a sheet. That is price monitoring and lead scraping on a flat plan, listed on the pricing page.
FAQ
Diffbot pricing questions buyers ask
Diffbot has a Free plan with 10,000 credits a month, Startup at $299 a month for 250,000 credits, Plus at $899 a month for 1,000,000 credits, and a custom Enterprise plan. Overage is $0.001 per credit on Startup and $0.0009 on Plus. Every paid plan includes all four APIs, and Crawl starts at Plus.
Extracting one page costs 1 credit, or 2 through the datacenter proxy. A Natural Language document costs 1 and a web search costs 1. Exporting or enhancing one Knowledge Graph entity costs 25, and a facet record or an enhance with refresh costs 100. Crawling to find pages costs 0 credits but is gated to Plus and up.
Yes. Free includes 10,000 credits a month across all four APIs with no credit card, limited to 5 requests a minute. That is enough for a few hundred pages a day if a developer writes the calls, but it has no Crawl and no scheduler, so you run and store everything yourself.
On Startup a page costs one credit, which is about $0.0012 or $1.20 per 1,000 pages, and $2.39 per 1,000 through the proxy. The credit price is low. The real cost is the plan floor of $299 and the developer time to schedule runs, store the JSON and map it to your columns.
For business teams that want rows in a sheet, yes. WebRobot Launch is $79 a month, or $39 billed yearly, for 10,000 records with daily schedules, and Scale is $249 for 100,000 records with hourly runs and logins. Diffbot stays cheaper only on its Free plan or when you need Knowledge Graph company data.
Diffbot bills credits for requests that are processed, and its documentation lists the actions that cost zero, such as spidering a site for links. Hitting the quota or the rate limit returns a 429 error instead of a charge, so check your run logs before a large batch so retries do not burn credits on pages that return nothing.
Rows in your sheet instead of credits in your bill
Describe the data, pick the schedule, and get named columns in Google Sheets, Excel or the API. Flat monthly plans from $79.