Home / AI extraction APIs / Diffbot
Diffbot pricing
100,000 AI extraction operations a month cost $299.00 on Diffbot's Startup plan, with room to grow to 250,000 AI extraction operations on the same plan.
Checked Oct 12, 2026 on Diffbot's own pricing page
- Startup
- Plus
Your workloadType an exact volume or copy a link to this jobEdit
published AI-extraction meter only
Fixed for this page; volume is the control.Every available option was priced in advance from published vendor rules. Changing a control only switches between those checked options; nothing is sent anywhere and no account is needed.
Every Diffbot plan
| Plan | Covers | Billed monthly | Billed yearly, per month | Included | Past the allowance | Concurrent |
|---|---|---|---|---|---|---|
| Startup fits your volume | 10K to 500K extraction operations | $299.00 | not offered | 250,000 credits | $1.00 per 1,000 credits | not published |
| Plus | 1M extraction operations | $899.00 | not offered | 1,000,000 credits | $0.90 per 1,000 credits | not published |
Your job
What Diffbot bills you for
What you pay
Diffbot is also sold as one of the web scraping APIs (100,000 JavaScript-rendered pages a month: no public price).
About this price and its sources
How Diffbot credits vary across extraction, crawl, language, and Knowledge Graph work, what is unclear, and who fits.
Cite this price
100,000 AI extraction operations a month cost $299.00 on Diffbot's Startup plan, with room to grow to 250,000 AI extraction operations on the same plan. Price read from https://www.diffbot.com/pricing on Oct 12, 2026. Scraping API Prices, https://scrapingapiprices.com/diffbot-pricing
How the bill builds up
- Base rate1 creditper page
- × JavaScript renderingNot requiredNot part of this workload
- × Premium proxyNot required
- ÷ Success rate1Failure billing is not published; this is an assumption
- = Billed units per operation1 creditFrom this bill’s calculation
- = per 1,000 operations$2.99Startup, billed monthly
Credits and multipliers
JavaScript rendering
Billed failures
Success rate
Concurrency
Free tier
Renews monthly The bill at your volume
100,000 extraction operations · $299.00
Read the chart as a table
| Volume | Bill | Offer |
|---|---|---|
| 10,000 | $299.00 | Startup |
| 25,000 | $299.00 | Startup |
| 50,000 | $299.00 | Startup |
| 100,000 | $299.00 | Startup |
| 250,000 | $299.00 | Startup |
| 500,000 | $549.00 | Startup |
| 1,000,000 | $899.00 | Plus |
Follow each billed unit
-
Successes you want
100,000 AI extraction operations
Your stated workload. Every vendor on this site is priced against this same number.
- ÷ 1
Attempts you are billed for
100,000 requests
Failure billing is not published. This workload assumes no failures, so the missing rule does not change the bill.
- × 1
What the meter counts
100,000 credits
One attempt is one billed unit for this workload.
- =
Cash per month
$299.00 /mo
Startup: the cheapest way to buy this volume, not just the first plan big enough.
Bars share one linear scale across the 3 counted stages, so a multiplier shows as length. Units differ by stage and are labelled. Cash is not on that scale.
100,000 AI extraction operations → 100,000 credits → Startup $299.00/mo = $2.99 per 1,000 operations
What makes it cost more
One page is not always one billed unit. These are Diffbot's published charges for each kind of work.
| Work | Costs | Per | Source |
|---|---|---|---|
| AI extraction | 1 credit | page | source |
Cheaper for this job
Same 100,000 AI extraction operations a month; retrieval, browser, proxy, and model charges are separate unless the vendor price includes them, ranked by what you would actually be billed. APIs without a full public price aren't ranked.
What Diffbot is
Diffbot provides automatic page extraction APIs, site crawling, natural-language processing, and access to a structured Knowledge Graph of public web entities. It is for teams that want normalized articles, products, organizations, people, or other entities rather than only raw HTML. AgentQL, Firecrawl, and Browse AI are the nearest named comparisons, but they emphasize semantic page queries, crawl-ready content for applications, or no-code robots. Diffbot’s distinguishing scope is the combination of page-type extraction and an existing web-scale entity graph. Sources: Diffbot pricing, Diffbot Extract, Diffbot Crawl, Diffbot Knowledge Graph, AgentQL, Firecrawl pricing, and Browse AI pricing, checked Jul 26, 2026.
How billing works
One credit is the common accounting unit, but products consume different numbers of credits. Extracting a page, extracting through a datacenter proxy, processing a text document, exporting a Knowledge Graph entity, and refreshing an entity do not have the same weight. Crawling for links itself does not consume credits, while each successfully extracted crawl page does. Overage is charged at the current plan’s per-credit rate, and self-serve allowances refresh monthly without rollover. The common buying mistake is to treat one credit as one page across the entire platform. That mapping applies only to selected operations. Sources: Diffbot credit documentation and Diffbot pricing, checked Jul 26, 2026.
What Diffbot does not publish clearly
- An annual self-serve price on the reviewed pricing page (Diffbot pricing, Jul 26, 2026).
- A direct Extract API failure rule separated from the general statement that credits are consumed when requests are made (Diffbot credit documentation, Jul 26, 2026).
- One cost-per-page figure that covers Extract, Crawl, Natural Language, and Knowledge Graph operations (Diffbot credit documentation, Jul 26, 2026).
- A self-serve crawl allowance on the lower tiers; public rate-limit documentation makes Crawl available only on the higher listed tier (Diffbot rate limits, Jul 26, 2026).
Who it suits
Diffbot suits teams that need normalized entity data, several extraction types, or a Knowledge Graph alongside page extraction. It is less suitable for a buyer seeking only raw page retrieval at one uniform success-only rate, or for a low-tier team that requires Crawl. The reason is that credits unify the invoice but not the workload: different operations have different weights, access limits, and success semantics. A useful procurement estimate must start with the exact Diffbot operation mix, not a generic page count. Sources: Diffbot pricing, credit documentation, and rate-limit documentation, checked Jul 26, 2026.