Skip to content

Alternatives

Four ways to get web data, compared plainly.

Most teams choose between a scraper subscription, a scraper marketplace, a single provider’s API or a unified API like PrimaryAPI. Each is the right call for someone. Here is how they differ, including where we are not the best fit.

See the comparisonHow routing works

Facts about named products are dated, read October 30, 2026.

When a source breaks

who fixes it

  1. Scraper subscriptionsYou switch providers yourself
  2. Scraper marketplacesYou switch scrapers yourself
  3. Single-provider APIsYou add a second vendor yourself
  4. PrimaryAPIThe next healthy provider answers, inside the same call
Typical behaviour for each approach; details in the comparison below.

The options

Four approaches, four trade-offs

None of these is wrong. They put the work, the risk and the bill in different places.

  1. Scraper subscriptions

    A monthly plan with a quota of requests or credits from one scraping provider. You send URLs or queries; it handles proxies and browsers.

    Pricing
    Monthly tiers with a request or credit quota.
    Good fit when
    Steady, high volume against a few sites, and a team that wants raw pages and full control of parsing.
    Watch for
    Paying for quota you do not use. Parsing and schema changes stay with you, and one provider’s bad day is yours too.
  2. Scraper marketplaces

    Community-built scrapers that you run on someone else’s infrastructure, paying for compute time or per result.

    Pricing
    Per run, per compute unit or per result, set scraper by scraper.
    Good fit when
    Niche or long-tail sites where someone has already built a scraper, and you are happy to pick and tune it yourself.
    Watch for
    Every scraper has its own inputs and output shape, and upkeep depends on its author. Retries and switching are on you.
  3. Single-provider APIs

    One vendor’s API for one kind of data: a search results API, an email verification service, or a platform’s official API.

    Pricing
    Per call or monthly plans, set by each vendor.
    Good fit when
    One data type at scale, a direct contract or SLA, or official access to a platform.
    Watch for
    A separate key, bill and response shape for every provider you add, and no fallback when that provider fails.
  4. Unified data APIs

    One key and one balance in front of many data APIs from many providers. PrimaryAPI is one of these.

    Pricing
    Per call from a prepaid balance, or credits within a monthly plan, depending on the product.
    Good fit when
    Many kinds of data, uneven volume, and AI agents that need to find and pay for data on their own.
    Watch for
    A layer between you and the source, with its own margin. Niche sources may not be covered yet; ask how failures and misses are billed.

Side by side

The criteria that change your bill

Typical behaviour for each approach, next to the facts for our own product. Individual products vary, so check the ones you are considering.

  • Pricing model

    Scraper subscriptions
    Monthly plan with a quota
    Scraper marketplaces
    Per run, compute or result
    Single-provider APIs
    Set by each vendor
    Unified data APIs
    Prepaid balance or plan credits
    PrimaryAPI
    Credits, charged per call. Pay as you go has no monthly fee or minimum; plans include credits every month.
  • What a failed call costs

    Scraper subscriptions
    Depends on the plan
    Scraper marketplaces
    Depends on the scraper
    Single-provider APIs
    Depends on the vendor
    Unified data APIs
    Varies; ask before you commit
    PrimaryAPI
    Nothing when every provider fails, or when we refuse the call before routing.
  • When a source breaks

    Scraper subscriptions
    You switch providers yourself
    Scraper marketplaces
    You switch scrapers yourself
    Single-provider APIs
    You add a second vendor yourself
    Unified data APIs
    Some route around it, some leave it to you
    PrimaryAPI
    Automatic failover to the next cheapest healthy provider, inside the same call.
  • Response shape

    Scraper subscriptions
    Raw pages or provider-specific JSON
    Scraper marketplaces
    Different for each scraper
    Single-provider APIs
    Different for each vendor
    Unified data APIs
    Normalised or per provider, by product
    PrimaryAPI
    One input and output schema per API, whichever provider answers.
  • Agent support

    Scraper subscriptions
    Varies by provider
    Scraper marketplaces
    Varies; some offer MCP
    Single-provider APIs
    Varies by vendor
    Unified data APIs
    Often MCP; sign-up flows vary
    PrimaryAPI
    Keyless discovery over MCP, SKILL.md, and self-signup with capped trial keys a person approves.
  • Coverage

    Scraper subscriptions
    Any site you can parse
    Scraper marketplaces
    Broad, quality varies
    Single-provider APIs
    One kind of data
    Unified data APIs
    Many kinds of data; depth varies
    PrimaryAPI
    260 APIs across 49 platforms today, added as each passes our checks.

The first four columns describe approaches, not specific products. The PrimaryAPI column describes this product as of October 30, 2026.

Honest limits

When we are not the right choice

Better you hear it here than after you have integrated.

  • You need data behind a login

    We serve public, logged-out data and licensed providers only. We never use anyone’s account, password or cookies.

  • You need a contractual uptime guarantee

    We publish measured uptime from our request log, but we do not offer a service-level agreement today.

  • You make one kind of call at very high, steady volume

    A direct contract with a single provider may cost less once your volume is large and predictable.

  • You need to make decisions about people

    Credit, employment, housing and insurance decisions need a consumer reporting agency. We are not one, and our terms forbid that use.

  • We don’t have the data yet

    Ask for it. Until it ships, a marketplace scraper or the source’s own API may get you there sooner.

Checklist

Eight questions to ask any data provider

Including us. The answers decide what you actually pay and what happens on a bad day.

  1. What does a failed call cost? And a “not found”?

    Our answer: Failed everywhere: nothing. Not found: free unless the API page says misses are billed.

  2. Can I see what a call costs before I make it?

    Our answer: Yes: every API page lists its price in credits, a free quote gives the price for your exact input, and the pricing page says what a credit costs on each plan.

  3. Is there a subscription, a minimum or an expiry on what I prepay?

    Our answer: Pay as you go has no monthly fee or minimum, and plans are optional. Bought credits never expire; credits included with a plan roll over for 3 months.

  4. What happens when the upstream source breaks?

    Our answer: The next cheapest healthy provider answers inside the same call.

  5. Can I cap spend per key, per agent and per call?

    Our answer: Yes: spend limits per key, trial caps for agents and maxCredits per call.

  6. Do they publish measured uptime, and how it is computed?

    Our answer: The network page shows figures from our request log and explains the method.

  7. Where does the data come from?

    Our answer: Public, logged-out sources and licensed providers. Each API page shows its live health.

  8. How long are my inputs and results kept?

    Our answer: Full results for 24 hours by default. Restricted-API inputs only as a hash.

Judge it on a real call.

Welcome credits are enough to try several APIs. If we are not the right fit, you have lost nothing but a few minutes.