Browserless Pricing: Browserless Cost per PDF on Every Plan, Units and Overages Explained

Browserless defines the unit in one sentence on its pricing page: "A Unit is a block of browser time of up to 30 seconds per browser connection. Longer-running automations use an extra Unit for every 30 seconds. Session-reconnects count as new browser connections and incur a fresh Unit charge." Its documentation adds the rounding rule: "Partial increments are rounded up, so a session open for 31 seconds consumes 2 units," and it applies to every connection method, including the REST APIs.

For PDF work that means the clock covers everything the browser does for that document: starting the connection, loading the page, fetching fonts and images, running any JavaScript that draws charts or fills tables, waiting for whatever selector or network state you asked it to wait for, and printing. An invoice template rendered from inline HTML typically finishes in a second or two and costs one unit. A dashboard that pulls data from three APIs and waits for the network to go quiet can run past 30 seconds and cost two.

Two details matter more than the headline rate. First, the default connection timeout is 30 seconds, and Browserless recommends setting your own timeout so a hung page cannot keep burning units. Second, reconnects are billed as new connections, so a retry loop that reconnects after a failure pays a fresh unit each time. Neither is a trap. Both are simply how time based billing behaves, and both are worth knowing before you estimate.

The pricing page opens on annual billing, so the $25, $140 and $350 most people quote are the annual rates. Flip the toggle to monthly and they become $35, $200 and $500, which is the "save 30%" the page advertises read the other way round: month to month costs about 40 percent more. Annual billing also buys extra concurrency, five more browsers on Prototyping, ten more on Starter and twenty more on Scale, which the monthly plans do not get.

Overage is billed in batches rather than at month end only. Browserless says overage fees are charged "whenever overages reach the dollar amount of the subscription the customer is on, when changing plans, or at the end of the month," with an email notice at most once an hour once you start accruing them. On the Free plan there is no overage: requests are refused once the 1,000 units are used.

At full use of each plan and one unit per PDF, a Browserless PDF costs $0.00125 on Prototyping, $0.00078 on Starter and $0.0007 on Scale, all on annual billing. Those are low numbers, and for fast pages they are real.

The number moves when renders get slow. Every PDF that crosses 30 seconds costs two units, every one that crosses 60 costs three. At 50,000 PDFs a month the difference between "all under 30 seconds" and "all between 31 and 60" is $85 against $140 on annual billing, and on monthly billing $95 against $195. You rarely get all of one or the other. The honest estimate is a blend, and the only way to know your blend is to measure it on your own pages.

What pushes a render over 30 seconds is almost never the PDF engine. It is the page: a web font served from a slow host, a chart library waiting for data, lazy loaded images below the fold that the print step forces to load, or a wait condition like network idle on a page that polls. Trimming those makes the page faster for human readers too, so it is often worth doing regardless of which vendor you choose. We walked through the same effect on a per minute meter in CloudConvert API pricing, where the step is 60 seconds instead of 30.

On yearly billing and fast pages, the two land within a few dollars of each other at most volumes, and the winner alternates. Browserless is cheaper at 1,000 PDFs a month (its free tier covers them), at 20,000 ($25 against $49.50) and between 100,000 and 180,000 ($140 against $149.50). SitePDF is cheaper at 10,000 ($14.50 against $25), at 50,000 ($49.50 against $85) and at 250,000 ($149.50 against $259). Once renders take longer than 30 seconds, every Browserless figure moves up and SitePDF's does not, because a render is counted once whatever it takes.

Month to month the picture shifts toward Browserless at the low end. Its $35 Prototyping plan covers 20,000 fast PDFs, while our monthly Starter API at $29 covers 10,000 and the next tier is $99. If your volume sits between 10,000 and 30,000 fast renders and you will not commit to a year, Browserless is the cheaper invoice.

Price per render is close enough that it should rarely decide this on its own. What differs is what comes back from the call:

• Browserless returns the PDF bytes. Your app stores them, names them, serves them and decides how long to keep them.
• SitePDF returns a hosted PDF link, and with archive=true keeps a timestamped copy of the source page under your account for 90 days on Starter API, one year on Growth and three years on Scale. That is the part teams usually end up building themselves for invoices, statements and compliance records, as covered in website archiving.
• The meter: time on Browserless, documents on SitePDF. A time meter rewards fast pages; a document meter makes the monthly bill a count you can read off your own database.

We sell a competing product, so here is the list we would want if we were choosing. Concurrency: Browserless Prototyping allows 10 browsers at once on annual billing, Starter 40 and Scale 100. Our tiers run 3, 10 and 30 concurrent renders. If you render large batches against a deadline, that ceiling matters more than any per unit price. Breadth: the same units pay for screenshots, scraping, crawling, full Puppeteer and Playwright sessions and an AI agent API. If PDFs are one job among several browser jobs, one Browserless bill is simpler than two vendors. Getting past bot protection: built in residential proxies and captcha solving exist for pages you do not own. SitePDF renders pages and HTML you control. Self hosting: Enterprise includes a commercial license to run Browserless on your own infrastructure, including air gapped deployments. Compliance paperwork: Browserless lists SOC 2 Type II, GDPR and HIPAA compliance and a DPA on its pricing page.

If those are your reasons, stay. If the job is turning your own templates or pages into documents you keep, our Browserless alternative page covers the switch in detail, and Browserbase pricing compares Browserless with Browserbase, Steel and Hyperbrowser on the same job for teams that want a remote browser anyway.

Use the 1,000 free units as a measuring instrument, not a trial. Render 200 to 300 of your real documents, not a sample invoice, through the /pdf endpoint with the same wait options you will use in production, and record the time each call takes. Three numbers come out of that: the share of renders under 30 seconds, the share between 30 and 60, and the share that time out.

Multiply your monthly volume by 1 x the first share + 2 x the second + 3 x anything slower, then read the result against the plan table above. Add 10 to 20 percent for retries if your pages depend on third party services. That is a far better forecast than dividing a plan price by its units.

Then put the same documents through the converter at the top of this page and compare the output side by side. If the layouts match and the volume sits where our tiers are cheaper, the document meter is simpler to budget. If they match and Browserless is cheaper at your volume and speed, it is a good product and the unit maths favors it. The broader field, including per page and per credit vendors, is priced on one table in PDF API pricing compared.