Browser-Hour Pricing: Examples & Companies

3 companies in the corpus Updated stub analysis
Definition

Browser-Hour Pricing is a billing unit where each hour of headless browser compute time is metered, common in web scraping and browser automation platforms.

Also known as: Browser Session PricingHeadless Browser Hour Pricing

What is it

Browser-hour pricing is a billing unit where each hour of headless browser compute time is metered, common in web scraping and browser automation platforms.

The browser-hour is the scraping equivalent of the GPU-hour: you pay for the time a real browser engine is running — rendering JavaScript, executing event handlers, and waiting on network responses. Browserbase surfaces the unit most explicitly because it operates a fleet of cloud-hosted Chromium instances; the cost it must cover is raw compute time, not a proxy like page counts or bytes transferred. Its plans bundle an hour pool (1 hour free, 100 on the $20/mo Developer plan, 500 on the $99/mo Startup plan) and then meter overage at $0.12/browser hr (Developer) or $0.10/browser hr (Startup). Browserbase notes a typical scrape runs under two minutes, so 100 hours is roughly 3,000 page-level tasks.

What distinguishes browser-hours from simpler per-request models is the treatment of rendering complexity. A static HTML page might return in under a second; a React SPA with lazy-loaded content can hold a browser session open for 10–30 seconds. Per-request models hide that variation; browser-hours expose it. This makes the unit better aligned with actual infrastructure cost, though it is harder for customers to forecast without running representative test workloads — the kind of usage-based metering trade-off that trades predictability for fairness.

Firecrawl and ZenRows both run real browser sessions for JavaScript-rendered pages but express the browser-hour cost through their primary unit rather than as a headline hour. Firecrawl’s Interact endpoint costs 2 credits per browser minute, and ZenRows meters its Scraping Browser at $0.09 per session-hour (billed in 30-second increments) drawn from a shared dollar balance. The three-company corpus therefore spans a spectrum: Browserbase (a direct browser-hour meter), ZenRows (an explicit per-session-hour line inside a shared balance), and Firecrawl (browser minutes priced as credits).

One hour of a running browser · priced three ways
One browser hour — from a direct meter to a credit bundle DIRECT METER $0.10–0.12 Browserbase · /browser hr after 100 / 500-hr pool SESSION-HOUR LINE $0.09 ZenRows · /session hr 30-sec increments, shared balance CREDIT BUNDLE 2 cr Firecrawl · /browser min ≈120 credits/hour, from pool ← UNIT IS EXPLICIT UNIT HIDDEN IN CREDITS →

How it works

Browser-hour billing starts a timer when a headless browser session opens and stops when the session closes. The platform converts session time to billing units — usually fractional hours or minutes — and aggregates across the customer’s account for the billing period.

CompanyBrowser billing unitPublished rateFree tier
BrowserbaseBrowser-hours (session compute time)$0.12/hr (Developer, after 100 hrs); $0.10/hr (Startup, after 500 hrs)Yes — 1 browser hour free
FirecrawlBrowser minutes, priced in credits2 credits/browser minute (Interact); Scrape/Crawl 1 credit/pageYes — 1,000 credits/month
ZenRowsSession-hours (Scraping Browser)$0.09/session-hour, 30-second increments, from a shared balanceYes — 14-day trial ($1 allowance)

The unit math for browser-hour pricing:

Cost = hours × hourly rate, where hours = Σ(session_duration in hours) across all sessions in the billing period.

Worked example (Browserbase Startup plan). A team runs 800 browser hours in a month. The $99/mo Startup plan includes 500 hours, leaving 300 overage hours at $0.10/hr = $30. Total browser-hour cost for the month is $99 + $30 = $129, before Search, Fetch, proxy, or model-token charges.

Concurrency is an important secondary dimension. Most platforms gate the number of simultaneous browser sessions by tier — a customer running 5 concurrent browsers accumulates 5× the hours of a single-session customer for the same wall-clock time. Browserbase exposes concurrent-session limits per plan (3 on Free, 25 on Developer, 100 on Startup, 250+ on Scale), making concurrency a capacity dimension alongside the raw hour count; ZenRows similarly scales concurrency from 5 on the trial to 400 on Business 3K. For Firecrawl, browser time folds into the credit ledger: an Interact run burns 2 credits per browser minute, so heavy browser automation quietly draws down the same monthly pool that a plain 1-credit-per-page Scrape uses. See billing-cycle mechanics for how these aggregated meters roll up into an invoice.


Companies using this

All three companies in the corpus meter browser compute time, differing only in how visible the hour is: a direct meter on Browserbase, an explicit session-hour line inside a shared balance on ZenRows, and browser minutes folded into credits on Firecrawl. Each targets engineering teams building web data-extraction and browser-automation pipelines at scale.


Company Product Pricing modelBilling unitsFree tier Verified
BrowserbaseBrowser-agent infrastructure: headless browser sessions, web Search/Fetch APIs, agent identity, runtime, and a model gateway behind one API keyYes2026-06-02
FirecrawlWeb-scraping and data-extraction API for AI agents — scrape, crawl, map, search, and extract pages into clean markdown/JSONYes2026-07-30
ZenRowsFetch, Extract, Batch, and Browser Sessions (plus legacy standalone Residential Proxies)Yes2026-08-04

Explore this theme in the knowledge graph

FAQ

What is browser-hour pricing for web scraping platforms?

Browser-hour pricing meters the amount of headless browser compute time consumed — each hour of a running Chrome or equivalent instance counts against a metered allowance. Browserbase charges per browser hour of session time (overage $0.12/hr on Developer, $0.10/hr on Startup); Firecrawl and ZenRows fold browser compute into their credit or shared-balance bundles. The unit is analogous to GPU-hours in ML infrastructure: it captures the dominant cost — compute while the browser renders JavaScript — rather than a proxy like page count.

How much does a browser hour cost?

On Browserbase, browser-hour overage is $0.12 per browser hour on the $20/mo Developer plan and $0.10 per browser hour on the $99/mo Startup plan, after each plan's included pool (100 and 500 hours). Firecrawl expresses browser time as credits — its Interact endpoint costs 2 credits per browser minute (≈120 credits/hour). ZenRows meters its Scraping Browser at $0.09 per session-hour, billed in 30-second increments, drawn from a shared dollar balance.

How does browser-hour pricing compare to per-request pricing for scraping?

Per-request pricing charges per URL fetched regardless of how long the page takes; browser-hour pricing charges for actual compute time. For heavy JavaScript-rendered sites a single page can hold a browser session open for 10–30 seconds, making per-request pricing unpredictable. Browserbase meters browser hours directly so slow crawls of complex SPAs pay proportional to the compute consumed. ZenRows and Firecrawl use request- and credit-based models but surface browser time as a separate line (a $0.09 session-hour and a 2-credit browser minute, respectively), bridging the two approaches.

Which companies use browser-hour pricing?

Three companies in the UsagePricing blueprint corpus meter browser compute time: Browserbase (a direct browser-hour meter), Firecrawl (browser minutes priced in credits via its Interact endpoint), and ZenRows (a $0.09/session-hour Scraping Browser drawn from a shared balance). All three target engineering teams building web data-extraction and browser-automation pipelines.

Related billing units

Back to companies