Pricing
FairCrawl uses cost-plus pricing. The price you see is tied to the route or worker method that actually ran.
Core prices
| Action | Typical billed price |
|---|---|
web/scrape | starts at $0.0007 per standard page; volume discounts apply |
web/crawl | based on the underlying page reads |
map | $0.0005 per request |
interact | $0.002 per request |
agent | $0.005 coordination cost before tool usage |
research | $0.007 per successful query, billed from the first query; synthesis included; successful enrichment pages separately metered at the canonical scrape rate |
pdf.fast | $0.00065 per page |
pdf.ocr | $0.0065 per page |
Research failures, queries with no results, and failed enrichment pages cost $0.
Method overrides matter
For web/scrape, the final bill can move depending on the worker method:
- direct fetch keeps the cheapest path
playwright/cf_browserescalations cost more than a plain static fetchpdf.fastandpdf.ocrbill per page- premium or proxy-backed fallbacks bill at the cost of the path actually used
Inspect pricing without guessing
Use the endpoint catalog as the public pricing map
No free tier
Every action bills cost-plus from the first call — nothing is bundled and no endpoint has an allowance. New accounts get a one-time $0.50 welcome credit in the wallet, which spends like money and runs out.
Work that delivers nothing still costs nothing: a failed, empty, or unscrapable
result settles at $0. That is not an allowance, it is us not charging you for
nothing.
The authenticated /v1/crawl/endpoints catalog is the source of truth for all
other action prices. A response’s _meta block records what that call cost.
Best practice
Map before you crawl, and scrape before you interact. That keeps your bill aligned with the cheapest endpoint that can answer the question.