Firecrawl scrape timeout: 60000 ms default vs Sume's 30000 cap

Firecrawl's scrape reference lists a timeout default of 60000 ms and a maximum of 300000 ms. Sume's scrape (POST /v1/firecrawl/scrape, MCP tool crawl_scrape) takes timeout_ms from 1 to 30000, default 30000. A request written for the longer Firecrawl window needs timeout_ms of 30000 or less here.

Firecrawl numbers are from its own pages; Sume numbers are from the OpenAPI document and the MCP tool description, all read 2026-09-30.

What are the two timeout ranges?

The field name differs as well as the range. Firecrawl calls it timeout; Sume's schema names timeout_ms and sets additionalProperties to false, so send the Sume name.

What request should I send?

Pass timeout_ms explicitly so the limit is visible in your code.

curl -X POST https://api.sume.com/v1/firecrawl/scrape \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com/pricing",
    "formats": ["markdown"],
    "timeout_ms": 30000
  }'

What happens when the page does not return in time?

The result is a typed timeout error rather than a retry loop; see MCP tools and gates for the tool list.

What if the page needs more than 30 seconds?

Scrape a more specific URL or ask for fewer formats. Related: the deadline layers for one sync scrape.