Get started

Pricing & credits

One balance of credits powers every API. Here is exactly what each call costs — and what it costs us to run, so you can see why ScrapeFlow is cheaper.

What a credit buys

Credits are the single unit across all endpoints. You are only charged on a successful 2xx response.

OperationCost
Scrape (1 format)1 credit
Scrape (each extra format)+1 credit
Map1 credit
Search2 credits
Brand lookup2 credits
Crawl1 credit / page
Batch1 credit / URL
Monitor1 credit / check
Answers (research)5 credits

The real unit economics

A single scrape flows through API Gateway → a router Lambda → SNS → SQS → a scraping Lambda → DynamoDB. Because it is all serverless and on-demand, we pay per request with zero idle cost. Here is the breakdown for one scrape:

ComponentPer scrapeCost
API Gateway (HTTP API)1 request$0.0000010
Router Lambda (128 MB, 80 ms)1 invoke$0.0000004
SNS publish1 message$0.0000005
SQS (send + receive + delete)3 requests$0.0000012
Scraper Lambda (1.5 GB, 3.5 s)1 invoke$0.0000877
DynamoDB write (on-demand)1 WCU$0.0000013
S3 (screenshot PUT + storage)1 object$0.0000050
Data transfer out (~50 KB result)egress$0.0000045
Infrastructure subtotalexcl. proxy~$0.0000966
ℹ
Proxy bandwidth is the real variable. Compute and messaging cost fractions of a cent; the dominant cost is egress through proxies. Direct fetches are effectively free, datacenter proxies add ~$0.0007/scrape, and residential proxies ~$0.0044. We route each request through the cheapest tier that works, then pass the savings on.

Cost calculator

Estimate your monthly cost, compare it against Context.dev list pricing, and see the serverless cost underneath.

1K1M

AWS cost to run

$9.66

$0.00010 / scrape

ScrapeFlow price (Pro)

$39.00

~75% gross margin

Context.dev price (Pro)

$99.00

list pricing

You save with ScrapeFlow: $60.00 / mo61% cheaper

Estimates use public AWS on-demand rates (us-east-1) and list prices. Actual proxy cost depends on page size and provider; direct fetches carry no bandwidth surcharge.

Why ScrapeFlow is cheaper

  • • No idle servers. Lambda scales to zero — we never bill you for a warm fleet waiting for traffic.
  • • Fan-out, not over-provisioning. SNS → SQS spreads bursts across ephemeral workers instead of a fixed cluster.
  • • Smart proxy routing. We only reach for residential proxies when a site demands it, keeping blended bandwidth low.
  • • On-demand storage. DynamoDB and S3 are pay-per-use, so cold data costs almost nothing.

Per-credit comparison

TierScrapeFlow / creditContext.dev / creditYou save
Developer$0.00036$0.00253~86%
Pro$0.00016$0.00079~80%
Growth$0.00013$0.00060~78%
Overage (Growth)$0.0003$0.0014~79%