Crawl4AI vs Zyte API

Crawl4AI 35

Graded on the hosted lines, for Crawl4AI Cloud, which the README and llms.txt now lead with and which an agent calls through its MCP server; the library would read on the local lines. No status page is linked, and status.crawl4ai.com fails TLS (0). No incident history readable (5). Per-plan limits are published with numbers in /v1/prices, from 10 requests a minute and 2 at a time on Free to 1,500 and 50 on Scale (15). 429 carries X-RateLimit headers and a message, 503 busy says to resend after a few seconds, and MCP job submissions take an idempotency key, but no Retry-After or backoff guidance (10 of 15). No SLA found in the terms or pricing (0). Prices and the free credit are a soft launch, /serp is beta and /answer experimental, while scrape, extract and search carry no beta label (5 of 10). 0 + 5 + 15 + 10 + 0 + 5 = 35.

Zyte API 85

Graded as a hosted API. status.zyte.com runs on Atlassian Statuspage with ten components and incident history (20). In the 90 days to 1 November 2026 its history feed lists two Zyte API incidents, both degraded browser rendering with higher latency and 520 errors, on 10 to 11 September and 13 September, neither marked as an outage, and three incidents outside the API, Scrapy Cloud periodic jobs on 17 August and the web dashboard and Scrapy Cloud periodic jobs down from 28 to 30 October, which Zyte says affected no other product. All partial, so minor (20 of 30). Rate limits are published as 3,000 requests a minute per key on standard plans and a custom limit on Enterprise, though per-site limits aren't (15). 429, 503 and 520 carry Retry-After, and the docs give a backoff table and say which responses to retry, per the 9 October check (15). Enterprise plans list service level agreements for support response times. No uptime SLA is published (5 of 10). Zyte API is generally available, and Zyte MCP carries no beta label on 1 November (10).

Crawl4AI 55

No OpenAPI or similar contract for the cloud API, and the MCP tool schemas are only in the live tools/list, which needs a key. /v1/prices is machine-readable but covers prices, not calls (0). llms.txt of 304 lines written for agents (10). Each endpoint says what it is for and when to choose another, such as trying without a proxy first, using browser only when auto misses content and /scrape/jobs for slow lists (16 of 20). Parameters are listed with types, defaults, allowed values (format both, md or html; proxy none, isp or residential) and limits, in prose (10 of 15). Curl examples per endpoint and documented error strings per status with whether each is charged (14 of 15). No changelog for the cloud API, only dated notes such as /serp 'added 8 Oct 2026'. The library has a dated CHANGELOG (5 of 15). 0 + 10 + 16 + 10 + 14 + 5 = 55.

Zyte API 85

The reference prints an OpenAPI 3.0.3 document for POST /v1/extract, though the Search API's /v1/search is documented in tables outside it (25). docs.zyte.com has an llms.txt and Markdown twins of its pages (10). The operation description states which output fields can be combined and which can't, and field descriptions give defaults (17 of 20). Inputs are typed, with 51 enums and required fields marked. customAttributes takes a free-form schema (13 of 15). The spec has a dozen request examples, and each error status has a problem+json example (15). The path is versioned /v1 and the spec says 1.0.0, but we found no dated changelog for the API in the docs sitemap of 150 pages (5 of 15).

Crawl4AI 46

Bearer keys, shown once and regenerable, with 3 to 30 keys per plan. claude.ai and ChatGPT sign in through a flow that advertises one scope, mcp, and each app shows as a revocable key. No per-key scopes, and the docs don't offer the key in a URL (22 of 30). A key can call billing routes, and the MCP topup and recharge tools charge the saved card, though llms.txt tells agents to ask first and accounts can set a spend cap. No read-only key (6 of 20). Scraped pages are untrusted text and no prompt-injection guidance was found in llms.txt or the API docs (0). Per-call cost, balance and fetch path in response headers and fields, no access log documented (6 of 15). The security page (4 October 2026) gives a disclosure process with fix targets and says SOC 2 is in progress, and the library publishes GitHub advisories with reporter credits. security.txt redirects to a 404, and no bounty found (12 of 20). 22 + 6 + 0 + 6 + 12 = 46.

Zyte API 55

API keys are revocable, an organisation can hold several, and none travels in a URL. Keys carry no permission scopes. Zyte MCP signs in over OAuth and works through its own key, mcp_access_key (22 of 30). An owner can set a blocking spending limit per key. We found no read-only mode, and the MCP docs warn that the agent can delete schedules and collections with no undo, leaving approval to the MCP client (8 of 20). The API returns untrusted web content, and the only guidance found is to review what the agent asks to do (2 of 15). The Stats API and dashboard break usage down by key, domain and request type, requests can be tagged, and each response has a request-id. No per-request log for the operator was found (10 of 15). ISO 27001 certification per the trust centre, a responsible disclosure programme with a hall of fame and no rewards, penetration test reports under NDA, and no security.txt (www.zyte.com/.well-known/security.txt answered 404 on 1 November) (13 of 20).

Crawl4AI 83

Library v0.9.5 on 14 October 2026, 23 days before 6 November, and the cloud added /serp on 8 October (30). v0.9.3 (31 August), v0.9.4 (23 September) and v0.9.5 (14 October) in the 90 days from 8 August (20). 56 open issues on 6 November, the newest triaged with Bug, Needs Triage, In-progress, Root caused or Done labels within days, plus a Discord. Reply authors weren't read (20 of 25). Not in the official MCP registry, and the cloud has no official SDK; the Python library is current on PyPI (8 of 15). The security workflow runs a test suite and a default-posture gate, and releases are built and checked with twine. No general test workflow on push found (5 of 10). 30 + 20 + 20 + 8 + 5 = 83.

Zyte API 77

The Agentic Web Data plugin (zytedata/skills) released 0.4.1 on 26 October 2026, 6 days before 1 November (30). Every dated source counted. In the 90 days from 3 August the plugin released 0.3.0, 0.3.1, 0.4.0 and 0.4.1 (17 and 22 September, 5 and 26 October) and scrapy-zyte-api 0.36.0 on 20 August. python-zyte-api 0.11.0 of 23 July has left the window, and its main branch holds unreleased work since (a CVE floor for brotli, Python 3.15 support), with no release by 1 November (20). There is no public API changelog. Support runs through tickets and a Discord server, and we did not read GitHub issues (8 of 15 for a closed service). python-zyte-api 0.11.0 is current, with Python 3.10 or later, and covers one language. The official MCP registry lists only a third-party Zyte entry (io.github.pipeworx-io/zyte), none from Zyte (12 of 15). The client repository has test and publish workflows and merged dependency updates on 20 and 26 October 2026. We did not see CI results (7 of 10).

Crawl4AI 67

16 MCP tools, six of them for billing and support (15 of 25). format md or html, parse opt-in blocks, batch limits and job results paged 500 at a time over HTTP and 50 over MCP (16 of 20). String error codes an agent can branch on (blocked, login-wall, no-answer, busy, no_credit, plan_cap, spend_cap), each with whether it is charged (18 of 20). Failed and cancelled calls aren't charged and jobs_submit takes an idempotency key, but no tool annotations are documented and two tools can charge the saved card (10 of 20). Only url or q is required. There is no official SDK for the cloud API; the docs give curl and standard-library Python and Node examples (8 of 15). 15 + 16 + 18 + 10 + 8 = 67.

Zyte API 72

Graded on the REST API, with the MCP server noted. A request names the outputs it wants, so an agent can ask for extracted fields or pageContent in place of full HTML, but HTML and bodies come whole, up to 10 MB. Zyte MCP lists 34 tools at once, 26 of them for Scrapy Cloud, recounted from its docs on 1 November (15 of 25). No pagination applies to a single-URL call. Search takes maxResults from 10 to 100, and the Stats API pages and filters (14 of 20). Errors are problem+json with a stable type path, and the docs say which to retry and which to fix (20). Calls are reads, failed and rate-limited requests are free, and retries are safe. We could not read the MCP tool annotations, which need a sign-in (14 of 20). A call needs only url and one output flag. The only official client is Python, with a Scrapy plugin (9 of 15).

Crawl4AI 45

No x402, MPP or L402 (0). Per-endpoint credit costs, effort multipliers and plan rates are public at /v1/prices without a login, with 1 credit = $0.001 (20). 1,000 free credits a month, no card (20). A 7-day pass key comes from a click on api.crawl4ai.com without an account, which is still a person in a browser, and a permanent key needs email verification (5 of 20). 0 + 20 + 20 + 5 = 45.

Zyte API 40

The official Python client documents x402 against api-x402.zyte.com, but our one unauthenticated request there on 1 November 2026, as on 9 October, returned 401 /auth/not-valid with WWW-Authenticate: Basic and no 402 challenge, and the main docs and pricing page don't mention it, so we score the line as not established (0 of 40). Per-1,000 prices for all five site tiers, screenshots and extraction are public, though a given site's tier needs the dashboard's cost estimator (20). $5 of credit for the first billing month on standard plans ($200 on Enterprise), with no card per the site, re-read in the pricing docs on 1 November (20). A person signs up in a browser, and the MCP route needs a browser sign-in too (0).