Free B2B AI Visibility Checker

Would an AI agent shortlist Firecrawl?

Last checked 2026-10-15 · 21 checks · Read-only, no browser

Results for firecrawl.dev

97out of 100

Shortlist-ready

AI agents can research Firecrawl, evaluate the product and get buyers to the sales team.

87%

Research

30.5/35 pts · 6/8 passed

93%

Evaluate

14/15 pts · 4/5 passed

100%

Reach out

43/43 pts · 6/6 passed

See what to fix first (3)

How a buyer's AI agent experiences Firecrawl

Research: passedEvaluate: passedReach out: passed

The agent read your site and understood what you do. It found your trial page at /signin?view=signup and could see how to reach your sales team.

What Firecrawl should fix first to make the shortlist

01

Company identity

Add Organization structured data (name, URL, logo, official profiles) to your homepage.

02

Page essentials

Give every key page a unique title, a one-sentence description, one main heading and a canonical URL.

03

Documented

Link developer docs from the homepage and publish your OpenAPI spec at a discoverable URL.

Research: can AI agents read and understand Firecrawl?

pass1.1

AI search crawler access

7/7 pts

AI search and assistant bots are allowed to read your site.

Training bots: GPTBot: allowed, ClaudeBot: allowed, Google-Extended: allowed, CCBot: allowed. Blocking these is a valid choice and doesn't affect this score.

Evidence
https://www.firecrawl.dev/robots.txt · HTTP 200 · blocked from /: none; blocked on key paths: none
pass1.2

Homepage access

7/7 pts

Your homepage lets in requests that identify as ChatGPT, Claude and Perplexity, and serves them the same page as a browser.

Evidence
https://www.firecrawl.dev/ · as browser · HTTP 200 · browser baseline 200
https://www.firecrawl.dev/ · as ChatGPT-User · HTTP 200 · HTTP 200
https://www.firecrawl.dev/ · as Claude-User · HTTP 200 · HTTP 200
https://www.firecrawl.dev/ · as Perplexity-User · HTTP 200 · HTTP 200
pass1.3

Content without JavaScript

7/7 pts

Your pages carry real text in the page code (3973 words on the homepage, 1092 on a typical key page).

Evidence
https://www.firecrawl.dev/ · homepage: pass (3973 words) · key pages: pass (median 1092 words across 5)
https://www.firecrawl.dev/ · 3973 words of page content in raw HTML (4095 including nav and footer)
sampled pages: product /use-cases/product-teams (2081 words), customer /blog/customer-story-stanford (905 words), blog /tools/article-summarizer (1092 words), about /about (582 words), product /tools/product-data-extractor (1116 words); median words: 1092
pass1.4

Accidentally hidden pages

4/4 pts

None of your key pages are hidden from search and AI.

Evidence
checked homepage + 5 key pages
warning1.5

Page essentials

1.5/3 pts

4 of 6 pages we checked are missing a canonical.

How to fix: Give every key page a unique title, a one-sentence description, one main heading and a canonical URL.

Evidence
https://www.firecrawl.dev/ · title:true desc:true h1:1 canonical:false
https://www.firecrawl.dev/use-cases/product-teams · title:true desc:true h1:1 canonical:false
https://www.firecrawl.dev/blog/customer-story-stanford · title:true desc:true h1:1 canonical:true
https://www.firecrawl.dev/tools/article-summarizer · title:true desc:true h1:1 canonical:false
https://www.firecrawl.dev/about · title:true desc:true h1:1 canonical:true
https://www.firecrawl.dev/tools/product-data-extractor · title:true desc:true h1:1 canonical:false
fail1.6

Company identity

0/3 pts

Your site doesn't identify the company in structured data.

How to fix: Add Organization structured data (name, URL, logo, official profiles) to your homepage.

Evidence
https://www.firecrawl.dev/ · 0 JSON-LD blocks, no Organization with a name
pass1.7

Sitemap

2/2 pts

Your sitemap is valid, linked and up to date.

Evidence
https://www.firecrawl.dev/sitemap.xml · valid:true linkedFromRobots:true urls:1020 latestLastmod:2026-10-15T04:37:31.401Z
pass1.8

Response speed

2/2 pts

Your homepage responds quickly (230 ms, median of 5 requests).

Evidence
https://firecrawl.dev/ · median 230 ms across 5 requests
https://firecrawl.dev/ · as crawler · HTTP 200 · 420 ms to response headers
https://www.firecrawl.dev/ · as browser · HTTP 200 · 230 ms to response headers
https://www.firecrawl.dev/ · as ChatGPT-User · HTTP 200 · 197 ms to response headers
https://www.firecrawl.dev/ · as Claude-User · HTTP 200 · 216 ms to response headers
https://www.firecrawl.dev/ · as Perplexity-User · HTTP 200 · 323 ms to response headers
bonus1.9

AI guide files

+1 bonus

You publish an llms.txt file. Early, and a good signal.

Evidence
https://www.firecrawl.dev/llms.txt · HTTP 200 · valid llms.txt (markdown, starts with # heading)
no <link rel="alternate" type="text/markdown">

Evaluate: can AI agents try Firecrawl's product?

pass2.1

Discoverable

4/4 pts

AI agents can discover your MCP server from your website, at https://mcp.firecrawl.dev/v2/mcp.

Evidence
https://www.firecrawl.dev/.well-known/mcp.json · HTTP 200 · valid JSON
https://www.firecrawl.dev/.well-known/mcp/server-card.json · HTTP 404 · HTTP 404
https://www.firecrawl.dev/.well-known/ai-catalog.json · HTTP 404 · HTTP 404
https://www.firecrawl.dev/.well-known/api-catalog · HTTP 404 · HTTP 404
https://www.firecrawl.dev/.well-known/oauth-protected-resource · HTTP 200 · valid JSON
https://www.firecrawl.dev/ · homepage mentions MCP: homepage text mentions an MCP server
pass2.2

Working

4/4 pts

Your MCP server responded in 459 ms.

Evidence
https://mcp.firecrawl.dev/v2/mcp · HTTP 200 · initialize: HTTP 200 in 459 ms
https://mcp.firecrawl.dev/v2/mcp · HTTP 200 · tools/list: HTTP 200
pass2.3

Understandable

3/3 pts

Your 3 tools are clearly described, so an agent can pick the right one.

Evidence
https://mcp.firecrawl.dev/v2/mcp · 3 tools, 3 with description >= 20 chars and inputSchema
pass2.4

Connectable

2/2 pts

An agent can work out how to connect to your MCP server.

Evidence
https://mcp.firecrawl.dev/v2/mcp · HTTP 200 · WWW-Authenticate: none
https://www.firecrawl.dev/.well-known/oauth-protected-resource · HTTP 200 · protected resource metadata found
warning2.5

Documented

1/2 pts

We couldn't find an OpenAPI spec.

How to fix: Link developer docs from the homepage and publish your OpenAPI spec at a discoverable URL.

Evidence
https://www.firecrawl.dev/ · docs link: https://docs.firecrawl.dev/
no OpenAPI found (api-catalog service-desc or homepage link)

Reach out: can an AI agent get a buyer to Firecrawl's sales team?

pass3A

Find the way in

8/8 pts

AI agents can find your trial page (/signin?view=signup) from the homepage.

Evidence
https://www.firecrawl.dev/signin?view=signup · trial page found via homepage
pass3B

Let in

10/10 pts

Your trial page lets in requests that identify as ChatGPT, Claude and Perplexity.

Evidence
https://www.firecrawl.dev/signin?view=signup · as browser · HTTP 200 · browser baseline 200
https://www.firecrawl.dev/signin?view=signup · as ChatGPT-User · HTTP 200 · HTTP 200
https://www.firecrawl.dev/signin?view=signup · as Claude-User · HTTP 200 · HTTP 200
https://www.firecrawl.dev/signin?view=signup · as Perplexity-User · HTTP 200 · HTTP 200
https://www.firecrawl.dev/robots.txt · robots disallow /signin?view=signup for: none
pass3C

See the form

12/12 pts

Your trial form is in the page code, so AI agents can see it.

Evidence
https://www.firecrawl.dev/signin?view=signup · as browser · HTTP 200 · contact form in page code with 2 fields (email, password)
pass3D

Understand the form

5/5 pts

Your form is clearly labelled with 0 required fields. An agent can fill it for its buyer.

Evidence
https://www.firecrawl.dev/signin?view=signup · 2 fields, 0 unlabelled, 0 required, email type: true
bonus3F

Agent-friendly route

+1 bonus

You give agents a direct route: visible email.

Evidence
email:[email protected] contactPoint:false llms.txt contact:false MCP contact tool:false
pass3G

Readable proof

5/5 pts

Your case studies are readable text, so agents can use them as proof.

Evidence
https://www.firecrawl.dev/blog/customer-story-stanford · HTTP 200 · 905 words of page content in raw HTML (nav and footer excluded)
pass3H

Self-serve path

3/3 pts

Agents can find and reach your free trial / sign-up.

Evidence
https://www.firecrawl.dev/signin?view=signup · as ChatGPT-User · HTTP 200 · 116 words, form: inline

What Firecrawl gets right

  • AI search and assistant bots are allowed to read your site.
  • Your homepage lets in requests that identify as ChatGPT, Claude and Perplexity, and serves them the same page as a browser.
  • Your pages carry real text in the page code (3973 words on the homepage, 1092 on a typical key page).
  • None of your key pages are hidden from search and AI.
  • Your sitemap is valid, linked and up to date.
  • Your homepage responds quickly (230 ms, median of 5 requests).
  • AI agents can discover your MCP server from your website, at https://mcp.firecrawl.dev/v2/mcp.
  • Your MCP server responded in 459 ms.
  • Your 3 tools are clearly described, so an agent can pick the right one.
  • An agent can work out how to connect to your MCP server.
  • AI agents can find your trial page (/signin?view=signup) from the homepage.
  • Your trial page lets in requests that identify as ChatGPT, Claude and Perplexity.
  • Your trial form is in the page code, so AI agents can see it.
  • Your form is clearly labelled with 0 required fields. An agent can fill it for its buyer.
  • Your case studies are readable text, so agents can use them as proof.
  • Agents can find and reach your free trial / sign-up.

Next steps

Compare with a competitor

Comparisons are private. Only people with the link can see them.

Fixed something?

How we checkedIs this your company? Request an update or removal

The check shows where agents drop off. The Assessment shows what it's costing you.

Our Pipeline Readiness Assessment is a complimentary 5-day diagnostic across paid efficiency, organic and AI search visibility, and pipeline data, with a 30-minute debrief from a specialist.