Skip to content

Choosing a Zyte API alternative: features, credits and plan fit

If you are evaluating a Zyte API alternative, the useful question is not 'which is cheaper' but 'which plan matches my workload'.

OUTPUT FORMATSHTMLJSONMarkdown

This page lays out the feature set, the credit rules that drive real cost, and a decision checklist you can apply to any scraping API.

What actually drives cost in a scraping API

Headline monthly price hides the variable that matters: how many credits each request consumes. A plan with more credits can still cost more per useful page if JavaScript rendering or premium proxies multiply the credit cost per call. Before comparing plans, model your own request mix.

Three levers usually dominate:

  • Rendering. Requests that need a headless browser cost more credits than plain HTML fetches.
  • Proxy class. Premium or residential proxies cost more than classic rotating proxies.
  • Retry behaviour. If failed or retried requests are billed, your effective cost per successful page rises with target difficulty.

A pay-per-success model changes the third lever. The original provider states that it charges only for successful extraction, with no additional costs for retries or errors, and that you do not pay for retries, failed attempts or traffic. That can make cost more predictable when success rates vary, but it does not make a hard target free. Whether a given plan actually bills this way is a contractual term to confirm before you buy.

Feature coverage to check before switching

Feature parity is rarely all-or-nothing. Rather than comparing marketing bullet lists, test each capability against your own hardest target and confirm it is present on the tier you intend to buy.

  • Proxy rotation. The original provider states that its system rotates proxies automatically, removing the need to manage pools or locations by hand. Ask any alternative whether rotation is automatic and whether you can still pin a location when a target requires it.
  • Managed headless browser. The original provider states that it runs the browser infrastructure so you do not maintain virtual machines yourself. Confirm who operates the browser layer on the plan you intend to buy.
  • CAPTCHA handling. The original provider states that stealth techniques minimise CAPTCHA triggers. Lower trigger rates do not guarantee zero challenges, so measure the challenge rate on your own targets.
  • Flexible modes. The original provider offers a fully managed endpoint for ease and a proxy mode for more control. Weigh those two trade-offs and confirm which modes your plan includes.
  • Extraction options. CSS or XPath rules for structured output, plus full-page screenshots.
  • Language coverage. Code examples in Python, Go, JavaScript and others.
  • Support tier. Priority support and a dedicated account manager are typically gated to higher plans.

If your workload depends on any single item above, confirm it is present on the plan tier you intend to buy, not only on the vendor's feature page.

Worked example: turning requests into credits

Suppose you need 200,000 successful page fetches per month, and 40% of them require JavaScript rendering while 10% require premium proxies with rendering enabled.

Using the supplied credit rules:

  • 120,000 plain requests at 1 credit each = 120,000 credits.
  • 80,000 JS-rendered requests at 5 credits each = 400,000 credits.
  • 20,000 premium-proxy + JS requests at 25 credits each = 500,000 credits.

Total: roughly 1,020,000 credits per month. That lands near the 1,000,000-credit tier, so the next tier up is the realistic fit once retries and growth are included. Run the same arithmetic against any alternative's per-request credit table before comparing monthly prices.

Note that credits are valid for one month and do not roll over, so unused capacity is not a buffer for a slow month.

How the plan tiers are structured

The plans fall into two groups: standard plans and enterprise plans. Within each group, tiers differ mainly in monthly API credits and concurrent requests. The original provider's published plan table marks priority email support, a dedicated account manager and team management as included from the Business tier upward; these are provider-published plan claims to confirm before you rely on them.

The plan selector below is the reference for prices, credit allowances and concurrency. Prices are exclusive of VAT. Treat the figures as a dated reference and confirm current pricing before purchase.

Decision checklist

Work through these in order; stop at the first item that fails.

  1. Does the API cover your hardest target? Test the page type that most often blocks you, not your easiest one.
  2. Is your request mix credit-heavy? If more than a third of requests need rendering or premium proxies, model credits before comparing prices.
  3. Do you need concurrency above the standard ceiling? Check the concurrent-request figure for the tier you can afford.
  4. Do you need support SLAs or team seats? These usually start at a higher tier.
  5. Is billing tied to success? Confirm whether retries and errors are billed.
  6. Does the credit window match your seasonality? Monthly-expiring credits penalise spiky workloads.

If all six pass, the remaining difference is usually documentation quality and integration effort.

Feature comparison at a glance

The table below compares the two APIs on request volume at the $99/month tier and on five capability rows. Statuses reflect the original provider's published claims and have not been independently verified; confirm against each vendor's current documentation.

Plan and credit reference

The full plan table follows, including standard and enterprise tiers. All prices are exclusive of VAT. Credit allowances are monthly and do not roll over. Each successful request starts at 1 credit; JavaScript rendering costs 5 credits, and premium proxies cost 10 or 25 credits depending on rendering.

Feature comparison
Original providerZyte API
Requests for $99/mo1,000,000~660,000
Natural-language extractionyesno
Table-to-JSON parsingyesno
Screenshot capture via APIyesno
JS scenario scripting via APIyesno
Flat-rate, credits-based pricingyesno

Provider-published statuses; not independently verified. Confirm current capabilities before relying on any row.

Plan and credit reference
PlanPrice /moAPI CreditsConcurrent requestsJS renderingRotating & Premium ProxiesGeotargetingScreenshots & Extraction RulesDedicated Scraping APIsPriority Email SupportDedicated Account ManagerTeam Management
Hobby$1975,00025IncludedIncludedIncludedIncludedIncludedNot includedNot includedNot included
Freelance$49250,00050IncludedIncludedIncludedIncludedIncludedNot includedNot includedNot included
Startup (Recommended)$991,000,000100IncludedIncludedIncludedIncludedIncludedNot includedNot includedNot included
Business$2493,000,000200IncludedIncludedIncludedIncludedIncludedIncludedIncludedIncluded
Business +$5998,000,000400IncludedIncludedIncludedIncludedIncludedIncludedIncludedIncluded
Enterprise 14$99914,000,000500IncludedIncludedIncludedIncludedIncludedIncludedIncludedIncluded
Enterprise 24$1,59924,000,000600IncludedIncludedIncludedIncludedIncludedIncludedIncludedIncluded
Enterprise 41 (Recommended)$2,39941,000,000700IncludedIncludedIncludedIncludedIncludedIncludedIncludedIncluded
Enterprise 69$3,59969,000,000800IncludedIncludedIncludedIncludedIncludedIncludedIncludedIncluded
Enterprise 120$5,799120,000,000900IncludedIncludedIncludedIncludedIncludedIncludedIncludedIncluded

All prices exclusive of VAT. Credits are valid for one month and do not roll over. Dated reference; confirm current pricing and support entitlements.

Frequently asked questions

How many credits does one request cost?

At minimum one credit per successful request. JavaScript rendering costs 5 credits. Premium proxies cost 10 or 25 credits depending on whether rendering is enabled.

Do unused credits carry over to the next month?

No. API credits are valid for one month and do not roll over.

Am I billed for retries and failed attempts?

The original provider states that it charges only for successful extraction, with no additional costs for retries or errors, and that you do not pay for retries, failed attempts or traffic. Whether a given plan bills this way is a contractual term to confirm before you buy.

What does 'concurrent requests' mean?

It is the number of requests that can be sent at the same time. Each tier lists its own ceiling, from 25 on the Hobby plan up to 900 on Enterprise 120.

Which countries can requests be routed through?

The original provider states 200+ countries and territories, with 40+ available on classic proxies. Treat this as a dated reference and confirm current coverage.