Firecrawl

firecrawl-rest-v1restprimary

78% coverage

Offload Valuejudged · weight 15 · Oct 5, 2026

This is genuine external work the model cannot perform by reasoning: live web search across the open index, geo-targeted/time-filtered retrieval, fetching and rendering pages, converting them to markdown/summary, query-relevant highlighting, and tool-contract discovery. It supplies fresh data the model does not have and would otherwise need many scrape-parse-clean steps to obtain. It falls just short of the top band because it is retrieval and transformation rather than an irreversible real-world side effect or truly proprietary/regulated data; index-based search is also approximable by other providers or raw HTTP fetching plus in-context cleanup, though at substantial token cost.

• https://docs.firecrawl.dev/api-reference/endpoint/search

  The search endpoint combines web search with Firecrawl’s scraping capabilities to return full page content for any query.

• https://docs.firecrawl.dev/api-reference/endpoint/search

  The search endpoint combines web search with Firecrawl’s scraping capabilities to return full page content for any query.

Interaction Costmeasured · weight 15 · Oct 5, 2026

Median over 2 successful run(s): 15480 tokens, 2 calls. Ratio to the best in "web-search" (3 surfaces measured, v0.2 scale). Cold, without documentation: 1/1 succeeded.

Error Recoverabilityprobed · weight 15 · Oct 5, 2026

unknown_endpoint → HTTP 502, clarity 25/100 (field named: no, allowed values: no, RFC 9457: no) · wrong_type → HTTP 400, clarity 75/100 (field named: yes, allowed values: yes, RFC 9457: no) · hallucinated_enum → HTTP 400, clarity 50/100 (field named: no, allowed values: yes, RFC 9457: no) · unknown_field → HTTP 400, clarity 70/100 (field named: yes, allowed values: yes, RFC 9457: no)

• https://api.firecrawl.dev/v1/search

  HTTP 403: {"success":false,"error":"Unfortunately, your IP address looks suspicious, so Firecrawl can't be used without an API key from here. Sign up for a free API key at https://firecrawl.dev/k/ffa7pwcbfrt6 f

• https://api.firecrawl.dev/v1/nonexistent_resource_xyz

  HTTP 502: Bad Gateway

• https://api.firecrawl.dev/v1/search

  HTTP 400: {"success":false,"error":"Invalid request body","details":[{"expected":"number","code":"invalid_type","path":["limit"],"message":"Invalid input: expected number, received string"}]}

• https://api.firecrawl.dev/v1/search

  HTTP 400: {"success":false,"error":"Invalid request body","details":[{"code":"invalid_value","values":["markdown","html","rawHtml","links","screenshot","screenshot@fullPage","extract","json","product","menu"],"

Doc Legibilitymeasured · weight 12 · Oct 5, 2026

text/markup ratio 4.0% (weight 30) · renders without JavaScript: yes (weight 25) · machine-readable spec: none found (weight 25) · llms.txt: https://docs.firecrawl.dev/llms.txt (weight 10) · 15 runnable example(s) (weight 10). Weighted average over 100 points of available sub-criteria.

• https://docs.firecrawl.dev/api-reference/endpoint/search
• https://docs.firecrawl.dev/llms.txt

Auth Frictionmeasured · weight 10 · Oct 5, 2026

Authorization: Bearer <key>, verified by a successful call. One header, no ceremony.

• https://api.firecrawl.dev/v1/search

  HTTP 403: {"success":false,"error":"Unfortunately, your IP address looks suspicious, so Firecrawl can't be used without an API key from here. Sign up for a free API key at https://firecrawl.dev/k/ffa7pwcbfrt6 f

• https://api.firecrawl.dev/v1/nonexistent_resource_xyz

  HTTP 502: Bad Gateway

• https://api.firecrawl.dev/v1/search

  HTTP 400: {"success":false,"error":"Invalid request body","details":[{"expected":"number","code":"invalid_type","path":["limit"],"message":"Invalid input: expected number, received string"}]}

• https://api.firecrawl.dev/v1/search

  HTTP 400: {"success":false,"error":"Invalid request body","details":[{"code":"invalid_value","values":["markdown","html","rawHtml","links","screenshot","screenshot@fullPage","extract","json","product","menu"],"

Safety & Reversibilityprobed · weight 10

If I retry after a timeout, do I charge the customer twice?

Payload Efficiencyprobed · weight 8

Can I ask for only the fields I need, or do I have to swallow 200 attributes to extract one?

Task Atomicitymeasured · weight 6 · Oct 5, 2026

Median over 2 successful run(s): 15480 tokens, 2 calls. Ratio to the best in "web-search" (3 surfaces measured, v0.2 scale). Cold, without documentation: 1/1 succeeded.

Prior Knowledge Coveragemeasured · weight 5 · Oct 5, 2026

Median over 2 successful run(s): 15480 tokens, 2 calls. Ratio to the best in "web-search" (3 surfaces measured, v0.2 scale). Cold, without documentation: 1/1 succeeded.

Async Compatibilityprobed · weight 2

If the operation is asynchronous, can I get the result without hosting a web server?

Tool-Native (MCP)binary · weight 2

Does the vendor publish an official, maintained MCP server?