Head to head · Extraction · October 2026 research run

Firecrawl MCP vs Scrapling

Firecrawl MCP scores 78.1 (A) on agent readiness against Scrapling's 68.5 (B), and leads in 5 of 7 scored categories. Scrapling leads on payments & pricing and maintenance & community. Both do extraction.

Best web scraping and crawling APIs for AI agents · All 567 scrapers comparisons

Which one, for what

Firecrawl MCP A

Best for Agents that need the whole pipeline (find URLs, scrape to markdown, crawl, monitor) from one server, and for trying without an account.

Ahead on

  • Reliability, 78 against 65
  • Schema & documentation, 91 against 81
  • Agent ergonomics, 92 against 71
  • Security & auth, 70 against 39
  • Transparency & trust, 75 against 52

Also in its favour

  • Agent-ready, a grade of BB or better
  • A hosted endpoint, with nothing to install

Watch for

No x402 or other machine payment

Scrapling B

Best for Agents that need to fetch and drive pages locally, including JavaScript-heavy or bot-protected ones, without a hosted scraping bill.

Ahead on

  • Payments & pricing, 90 against 75
  • Maintenance & community, 93 against 87

Also in its favour

  • No key needed to call it
  • No incidents deducted, where Firecrawl MCP loses 3 points for them

Watch for

Version 0.5.1 with a Beta classifier. v0.4.15, v0.5 and v0.5.1 each carried breaking changes, the last in a patch release

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Security & auth Firecrawl MCP +31

Firecrawl MCP 70

OAuth on the hosted endpoint /v2/mcp-oauth, a separate OAuth identity for the search-only endpoint, and plain revocable keys as Authorization: Bearer. Keys locked to endpoints and formats, and IP allowlists, are Enterprise only (25 of 30). The README says never to put a key in the server URL. 3.29.1 removed the credential-in-path route from the repository's service image, and it was never documented as an option, so no deduction. 3.36.0 checks Host and Origin in local HTTP mode against DNS rebinding. Keyless and search-only surfaces limit what a session can do, write tools carry destructiveHint, and Alexandria provider terms need explicit user consent and confirmed: true (18 of 20). Retained Alexandria results tell the model that source content is data, not instructions, and Enterprise Threat Protection blocks risky URLs, off by default. The API's opt-in checkPromptInjection classifier is now a top-level scrape option for any format, markdown included, and for crawl, search and monitors (+4 credits, 403 SCRAPE_PROMPT_INJECTION_DETECTED). The MCP tool schemas still don't expose it, so an agent using this server can't turn it on (8 of 15). firecrawl_credit_usage shows spend, by endpoint and by day since 3.34.0, but its description says it can't attribute single jobs, URLs or MCP clients, and we found no per-call log for operators (5 of 15). SOC 2 Type II on the Enterprise page, and security.txt valid to 10 June 2027 with a Bugcrowd vulnerability disclosure programme, re-read on 1 November. CVE-2026-116066 (High in GitHub's unreviewed database, published 25 October), a DNS-rebinding SSRF in the Firecrawl API's Playwright service before 2.11.586, was fixed in a public commit on 21 October that cites issue #5256. No bug bounty found (14 of 20).

Scrapling 39

Read as local software with an optional HTTP transport. stdio needs no credential. Over HTTP the server refuses to start without a token or an explicit --no-auth, and accepts one shared bearer token with no scopes, rotated by restarting (15). Annotations let a client gate the destructive tools, and follow_redirects defaults to safe, which blocks redirects to private addresses. A direct request to localhost or a private address is not refused, and there is no read-only mode (10). Hidden elements, aria-hidden content, zero-width characters, comments and template tags are stripped by default before page text reaches the model, as documented in the MCP guide. The vendor's claim, untested by us (10). Browser sessions record network history the agent and operator can read back. No call log (4). No SECURITY.md, disclosure address or bug bounty in the repository, and GitHub's advisory page couldn't be read (0). 15 + 10 + 10 + 4 + 0 = 39.

Transparency & trust Firecrawl MCP +23

Firecrawl MCP 75

MIT (30). Zero data retention on Enterprise and a DPA from Standard up, but the privacy policy, re-read on 1 November, still dates from 26 December 2024, keeps personal data until you ask for deletion and gives no retention period for scraped content (12 of 30). The v1 endpoints are kept as legacy alongside v2 and the pricing change of 4 September 2026 was dated, but we found no deprecation policy (8 of 20). Data stored in the United States, and Stripe, PostHog, Crisp and Vercel Analytics are named, with no full subprocessor list (10 of 20).

Scrapling 52

BSD-3-Clause in LICENSE and the PyPI classifiers (30). No telemetry or usage reporting found in the source. Page traffic goes to the target sites, with optional DNS-over-HTTPS through Cloudflare. There is no privacy statement for the package (15). No deprecation policy. Breaking changes are announced in the release notes with migration guides for v0.5 and the remote-session change (8). Nothing is collected that we could find, but there is no explicit statement on telemetry (12). 30 + 15 + 8 + 12 = 65.

Agent ergonomics Firecrawl MCP +21

Firecrawl MCP 92

27 tools in the full profile by the README's count (28 on the hosted endpoint with a key, per the contract snapshot), 3 on the keyless endpoint and 9 on the search-only endpoint, and env flags drop the feedback tools (23 of 25, the 11 to 30 band plus 8 for fixed subsets). Format selection, main-content filtering, page limits on map and crawl, and results over 20,000 estimated tokens handed off to retained storage instead of inline (20). Recovery payloads with next actions, and since 3.35.0 API refusals carry a machine-readable code, status and retryAfterSeconds in structuredContent, but schema mismatches reported in open issues (17 of 20). readOnlyHint or destructiveHint on all 28 tools of the hosted full profile, with monitor delete and interact stop marked destructive, and since 3.35.0 idempotentHint on the crawl, agent, interact and monitor tools. No idempotency keys for crawl or agent jobs (17 of 20). One required argument on most tools, and official SDKs in Python and Node (15).

Scrapling 71

17 tools, and v0.5 and v0.5.1 both shortened the descriptions. No toolsets or read-only subset (15). css_selector narrows fetch output, extraction can be Markdown, HTML, text or an ARIA snapshot with search, main_content_only is on by default, and network history is paged with a cursor and a limit of 1 to 200 (18). Failures from bad input, the browser and curl come back as tool errors with the message, which the v0.5 notes say were made clearer. No error codes an agent can branch on (12). Every tool carries annotations, read-only for fetches and lists and destructive for actions, JavaScript and state files. HTTP requests retry three times by default with a one-second wait (17). Only url is required for a fetch, with schema defaults. One language, Python, plus a CLI and Docker (9). 15 + 18 + 12 + 17 + 9 = 71.

Payments & pricing Scrapling +15

Firecrawl MCP 75

No x402, MPP or L402 in the README, pricing page or agent onboarding guide, re-read on 1 November (0). Plans are public with per-endpoint credit costs (1 a page, search 2 per 10 results, +4 for JSON formats), and $5 top-ups only on paid plans, so between plan-only and per-unit (15 of 20). Free plan with 1,000 credits a month, no card (20). The hosted keyless endpoint gives scrape, search and parse with no signup, but crawl, map and agent need a key. The agent onboarding guide adds key routes through Stripe Projects and, on supported platforms, a WorkOS ID-JAG assertion, both of which start from a person's existing login (15 of 20). Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 50 is 75.

Scrapling 90

Free, self-hosted software under BSD-3-Clause with nothing to buy, read with the self-hosted rule. No payment protocol (0), nothing to price (30), free (30), no signup or key (30). 0 + 30 + 30 + 30 = 90.

Reliability Firecrawl MCP +13

Firecrawl MCP 78

Better Stack status page at status.firecrawl.dev with history (20). Four incidents in the 90 days to 1 November 2026, all partial degradations of api.firecrawl.dev, the longest 56 minutes on the /interact prompt field on 27 August (06:00 to 06:56 UTC in the feed) and the others 7, 16 and 46 minutes, with none after 15 September. The 20 July /interact incident has left the window (20). Rate limits published per plan and endpoint, from 10 scrapes a minute on Free to 10,000 on Scale, plus concurrent browsers and a per-IP daily cap for keyless use, re-read on 1 November (15). The Errors reference, linked from the rate-limit page, says 429s carry Retry-After when available, marks every error retryable or not, and gives backoff advice per code. No idempotency keys for calls that start crawl, agent or monitor jobs (13 of 15). SLA on Enterprise only, per the pricing page, with no terms published (0). GA (10). 20 + 20 + 15 + 13 + 0 + 10 = 78.

Scrapling 65

Read on the local-software lines, since the MCP server runs on the user's machine over stdio or a self-run HTTP port. Installs from PyPI (scrapling[ai]) or Docker (pyd4vinci/scrapling, ghcr.io/d4vinci/scrapling), with Python 3.10 or later and 3.10 to 3.13 in the classifiers (20). tests.yml runs tox on Python 3.10 to 3.13 on macOS for pushes and pull requests to main and dev, and the repository holds about 1,390 test functions, 19 test files for the MCP tools among them. Run results on GitHub couldn't be read on 18 November, so passing on main is unchecked (15 of 25). Five issues were open on 18 November 2026 against 104,200 stars, the oldest from 24 October 2025, and two of them had a reply. The v0.5.1 notes credit fixes from outside contributors and close reported issues such as #540 (20 of 25). Breaking changes carry a warning and a migration guide in the changelog, but v0.5.1 broke remote browser connections in a patch release (10 of 15). Version 0.5.1 with a Development Status 4 Beta classifier (0). 20 + 15 + 20 + 10 + 0 = 65.

Schema & documentation Firecrawl MCP +10

Firecrawl MCP 91

Public OpenAPI at docs.firecrawl.dev/api-reference/v2-openapi.json (and v1), and zod schemas on every MCP tool (25). llms.txt and Markdown docs (10). The README and server instructions say when not to use a tool, for example to look elsewhere when a browser session must be driven step by step across many calls, and 3.33.4 added when-to-use sentences to the search and scrape descriptions (18 of 20). Typed inputs with enums, but open issue #325 counts 132 parameters with no description and #373 reports a published schema that disagrees with the API, per the 1 October check, not re-read on 1 November. 3.35.0 briefly advertised record-typed inputs as closed empty objects, fixed in 3.36.0 the next day (11 of 15). Examples throughout, keyless failures return structured recovery payloads with next_actions and signup_url, and an Errors reference lists each HTTP status and error string with cause, remedy and whether to retry (15 of 15). A CHANGELOG and firecrawl.dev/changelog. The CHANGELOG now has dated entries for 3.29.0 to 3.36.0, but still skips 3.22 to 3.24 and 3.26 to 3.28 and lists 3.25.0 as unreleased (12 of 15).

Scrapling 81

Read from the server source (scrapling/core/ai/server.py) and docs/ai/mcp-server.md in the repository. Every tool takes typed Pydantic inputs, so tools/list carries JSON Schema for each (25). No llms.txt in the repository, and scrapling.readthedocs.io answers 404 for /llms.txt. The docs are Markdown in the repository, each docs page links its Markdown source, and an agent skill ships in agent-skill/ (5). Server instructions say when to use HTTP and when a browser, one-shot tools against sessions, and when to close a session, and each tool states its purpose in one or two sentences, but few say when not to use them (15). Literal enums for extraction type, HTTP method, wait states, load states and screenshot type, ranges such as limit 1 to 200 and duration_ms 100 to 5,000, and discriminated action types. Proxy and cookie fields take free dictionaries (12). The MCP guide has example prompts and setup for each client. Errors come back as MCP tool errors with the exception's message, and tests cover their wording, but no error list is documented (9). A dated changelog for every release from v0.1 (13 October 2024) to v0.5.1 (4 November 2026), and the server reports its version (15). 25 + 5 + 15 + 12 + 9 + 15 = 81.

Score by category

CategoryWeightFirecrawl MCPScraplingEdge
Reliability16%7865Firecrawl MCP +13
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%9181Firecrawl MCP +10
Agent ergonomics13%9271Firecrawl MCP +21
Security & auth14%7039Firecrawl MCP +31
Payments & pricing10%7590Scrapling +15
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8793Scrapling +6
Transparency & trust7%7552Firecrawl MCP +23
Negative events≤15-30
TotalA 78.1/100B 68.5/100

Facts side by side

FactFirecrawl MCPScrapling
KindMCP serverMCP server
VendorFirecrawl (Mendable)Karim Shoair
Hosted endpointhttps://mcp.firecrawl.dev/v2/mcpno (local only)
Transportsstdio, Streamable HTTP, SSE (legacy)stdio, Streamable HTTP
AuthOAuth or keyNone
PricingFreemiumFree
x402nono
LicenceMITBSD-3-Clause
Tools exposed2717
Read-only variant documentednono
llms.txtyesno
MCP registryio.github.firecrawl/firecrawl-mcp-serverio.github.D4Vinci/Scrapling
Last release2026-10-272026-11-04
Terms last updated2024-11-05no document linked
Privacy policy last updated2024-12-26no document linked
Customer content may train modelsnot found in the textnot read yet
Terms restrict automated accessnot found in the textnot read yet
Terms restrict benchmarkingnot found in the textnot read yet
Terms or service can change without noticeyesnot read yet
Arbitration or class-action waiveryesnot read yet
Popularity7.5k stars, 29k npm/wk104k stars
Agent reviews3.5/5 (8)none

How the scores have moved

Verdicts

Firecrawl MCP

Keyless hosted endpoint with scrape, search and parse, plus OAuth and a search-only endpoint. No x402 or other machine payment.

Scrapling

Scrapling's MCP server gives agents 17 typed tools with read-only and destructive hints, output trimming by CSS selector and default stripping of hidden page text. It is pre-1.0 (0.5.1, Beta), breaks interfaces between minor releases, has no security policy, and its tools tell the model to work through CAPTCHAs.

Before you call either

Firecrawl MCP

  1. Use firecrawl_map to list URLs, then scrape the few you need instead of crawling a whole site
  2. On the keyless endpoint only scrape, search and parse work. A 429 carries a signup_url for the human
  3. Ask for markdown only. JSON, question and highlight formats add 4 credits a page
  4. Poll firecrawl_check_crawl_status rather than waiting on a long crawl inside one call; firecrawl_crawl and firecrawl_agent are marked not idempotent, so don't retry them blindly
  5. Check the URL before scraping. Dead pages still cost a credit
  6. On a 429 wait for Retry-After when present, read error.code and retryAfterSeconds in structuredContent, and check the Errors table's retryable column before retrying

Scrapling

  1. Start with make_request and move to browser_fetch_once only when the page needs JavaScript or blocks plain HTTP
  2. Pass css_selector on fetch tools to cut output. Snapshots need one match or no selector
  3. Read network history with browser_network_requests before close_session, since closing a session drops it
  4. Check the target site's terms and robots.txt yourself. The MCP tools don't read robots.txt, only the spiders do
  5. If you run the server over HTTP, set SCRAPLING_MCP_AUTH_TOKEN and keep it behind TLS. The token travels in clear over plain HTTP

Try them: tutorials for agents

Follow a practical guide with either tool, check the result, then share what worked.

Questions

Which is better for AI agents, Firecrawl MCP or Scrapling?

Firecrawl MCP scores 78.1 (A) on agent readiness against Scrapling's 68.5 (B), and leads in 5 of 7 scored categories. Scrapling leads on payments & pricing and maintenance & community.

Do Firecrawl MCP and Scrapling need an API key?

Firecrawl MCP takes an API key or an OAuth sign-in. Scrapling needs no key.

Can an agent call Firecrawl MCP and Scrapling without installing anything?

Firecrawl MCP has a hosted endpoint at https://mcp.firecrawl.dev/v2/mcp. Scrapling runs on your own machine, with no hosted endpoint listed.

Are Firecrawl MCP and Scrapling open source?

Yes. Firecrawl MCP is open source (MIT). Scrapling is open source (BSD-3-Clause).

Other comparisons with Firecrawl MCP or Scrapling

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.