9 min read
Zyte API costs $0.13 to $1.27 per 1,000 successful HTTP requests and $1.01 to $16.08 per 1,000 browser-rendered requests on pay as you go. Where you land in that range is not your choice: Zyte puts every website into one of five difficulty tiers. A monthly commitment of $100 to $500 takes 25 to 52 percent off every tier, and failed or rate-limited requests are never billed.
figures taken from zyte.com/pricing and the Zyte API pricing documentation
Robot console
Standing by Running · s Complete · rows
1 · Pick a target
3 · Fields to extract
Agent log
Crawl graph
Extracted data · rows
Want this data fresh every morning, without lifting a finger?
The numbers
| Tier | Pay as you go | $100/mo commitment | $200/mo commitment | $500/mo commitment |
|---|---|---|---|---|
| HTTP response body, per 1,000 successful requests | ||||
| 1 · Simple | $0.13 | $0.10 | $0.08 | $0.06 |
| 2 · Easy | $0.23 | $0.17 | $0.14 | $0.11 |
| 3 · Moderate | $0.44 | $0.33 | $0.26 | $0.21 |
| 4 · Complex | $0.70 | $0.53 | $0.42 | $0.34 |
| 5 · Advanced | $1.27 | $0.95 | $0.76 | $0.61 |
| Browser rendered, per 1,000 successful requests | ||||
| 1 · Simple | $1.01 | $0.75 | $0.60 | $0.48 |
| 2 · Easy | $2.01 | $1.50 | $1.20 | $0.96 |
| 3 · Moderate | $4.02 | $3.00 | $2.40 | $1.92 |
| 4 · Complex | $8.04 | $6.00 | $4.80 | $3.84 |
| 5 · Advanced | $16.08 | $12.00 | $9.60 | $7.68 |
Two things jump out. First, rendering costs eight to thirteen times an HTTP request on the same tier, so whether a site needs a browser matters more than which tier it sits in. Second, the tiers double: a Complex site costs about five times a Simple one over HTTP and eight times with a browser. A job that looks cheap at the headline "$0.13" can easily be a $8 per 1,000 job once Zyte classifies the target.
The meter
| Monthly commitment | Discount on every tier | What it means in practice |
|---|---|---|
| None, pay as you go | 0 percent | List prices; the account is capped at $100 of spend a month |
| $100 | 25 percent | You pay at least $100 even in a quiet month |
| $200 | 40 percent | The step most production scrapers settle on |
| $350 | 48 percent | Only worth it once you reliably spend above the $200 level |
| $500 | 52 percent | Top self-serve step; Enterprise negotiates beyond it |
Billed
Screenshots cost $0.002 each. Automatic extraction, where Zyte's model returns a product, article or job posting as fields, adds $0.0004 to $0.0016 per data type. Custom attributes cost $0.002 per 1,000 input tokens and $0.01 per 1,000 output tokens with the generate method, or a fixed $0.001 with extract.
Not billed
Unsuccessful and rate-limited responses are free. Above your commitment you keep paying your discounted rate with no overage penalty. New accounts get $5 of credit for 30 days, and Enterprise contracts start with $200 of credit.
Proxies
Smart Proxy Manager, the product formerly known as Crawlera, is retired. Zyte moved the remaining traffic onto Zyte API proxy mode by December 2025, and existing plans became migration plans with the same monthly commitment and standard discounts. Proxy mode lets old Scrapy and requests code keep sending traffic through a proxy address, but the bill follows the same per-website tiers in the table above. You are paying for a successful response from a site of a given difficulty, not for bandwidth, residential IPs or sessions.
For most teams that is a better deal than a gigabyte meter, because blocked attempts cost nothing. It does make the monthly total harder to predict before you run the job, since the tier is assigned by Zyte after it sees the site. Paste your real target URLs into the calculator on the pricing page before you commit.
Real months
Worked bill 01
HTTP on tier 2 is $0.23 per 1,000, so $11.50 a month on pay as you go. Turn on automatic product extraction and each page adds up to $0.0016, which is as much as $80 more. Still under $100, and very cheap for 50,000 fetches.
Worked bill 02
Tier 4 with a browser is $8.04 per 1,000, or $241.20, which is over the $100 pay as you go cap. A $100 commitment drops the rate to $6.00: $180. A $200 commitment would be $144 of usage but you pay the $200 floor, so the smaller commitment wins.
Worked bill 03
About 9,000 rendered requests a month on a Moderate site at $4.02 is $36.18. Cheaper than WebRobot Launch at $79 on the invoice. The difference is the spider, the parser, the scheduler, the history and the Slack alert, which Zyte leaves to you.
The other half
Not included 01
Zyte API returns a page. Which pages to request, how to paginate and how to turn HTML into columns is a Scrapy project someone writes in Python, unless automatic extraction happens to cover your data type.
Not included 02
Spiders need somewhere to run on a timetable. Zyte sells Scrapy Cloud separately for that, or you run them on your own servers and cron. Either way it is another line in the budget.
Not included 03
A redesign breaks selectors and the spider starts returning empty fields. Ban handling does not fix parsing. Somebody on the team owns that fix every time.
Not included 04
Getting rows into a Google Sheet, a warehouse table or a Slack message when a price moves is code you build and host after the response arrives.
Teams that run their own pipeline on Zyte also end up watching it. A spider that silently stops is the classic failure, and it is worth putting uptime checks on the endpoints your pipeline depends on next to the scraping budget, so a dead job pages someone the same hour rather than the week the report comes out empty.
Comparison
| Tool | What you pay for | Entry price | What you still build |
|---|---|---|---|
| Zyte API | Successful responses, priced by site tier | Pay as you go from $0.13 per 1,000 | Spider, parser, hosting, schedule, delivery |
| Firecrawl | Credits: 1 per page, 5 with JSON extraction | Hobby $19/mo for 5,000 credits | Schema, schedule, delivery |
| Apify | Compute units, proxy traffic and per-result fees | Starter $19/mo plus usage | Actor choice or code, delivery |
| Bright Data | Records from its scraper APIs | $1.50 per 1,000 records pay as you go | Schedule and delivery |
| WebRobot | Records delivered by a hosted robot | Launch $79/mo for 10,000 records, $63 billed yearly | Nothing: one sentence describes the data, the robot schedules and delivers |
If you are weighing a move, our Zyte alternative comparison covers features side by side. The same fetch-versus-finished-data trade-off runs through Firecrawl pricing and Apify pricing, and the Bright Data alternative page covers the heaviest anti-bot option. For the whole market, see the cost of web scraping.
Decision
You have Python developers, you already run Scrapy, and your problem is bans on hard targets at hundreds of thousands of pages a month. Paying only for successful responses at $0.06 to $0.61 per 1,000 on a $500 commitment is excellent value, and nothing record-priced beats it at that scale.
The request comes from pricing, sales or operations and nobody wants to own a spider. Describe the data in a sentence, pick daily or hourly, and rows arrive in Sheets, Slack or your own system through the web scraping API. Typical jobs are competitor price monitoring and ecommerce data scraping. Plans are on the pricing page.
FAQ
Describe the data in plain English. The robot runs on a schedule and delivers 10,000 records a month to Sheets, Slack or your API for $79.