Benchmark report

Not supported

0m 17s elapsed

Official documentation shows this product does not offer the task. Tavily's documented scraping capability is the synchronous POST /extract endpoint: it accepts one URL or a batch of up to 20 URLs and returns results/failed_results plus response_time in the same HTTP response, with no job id or status to poll. The complete API reference index (llms.txt) lists only Search, Extract, Crawl, Map, Research (create/get/stream), Feedback, Usage, Logs and enterprise key endpoints — the only create-then-poll job endpoints are for the Research agent (cited report generation), not for batch URL scraping. There is therefore no documented asynchronous batch scrape/job API returning a job id with terminal status and per-URL results; the synchronous multi-URL extract call is the documented gap, and the task forbids using it as a substitute.

Where the time went

0:18 agent time
  • Research0:17
0m 17sTotal elapsed
—Time to first success
0:00Waited on the owner

Attempt record

  1. —Not supported by this productTavily's documented scraping capability is the synchronous POST /extract endpoint: it accepts one URL or a batch of up to 20 URLs and returns results/failed_results plus response_time in the same HTTP response, with no job id or status to poll. The complete API reference index (llms.txt) lists only Search, Extract, Crawl, Map, Research (create/get/stream), Feedback, Usage, Logs and enterprise key endpoints — the only create-then-poll job endpoints are for the Research agent (cited report generation), not for batch URL scraping. There is therefore no documented asynchronous batch scrape/job API returning a job id with terminal status and per-URL results; the synchronous multi-URL extract call is the documented gap, and the task forbids using it as a substitute.Tavily Extract API ReferenceTavily Docs index (llms.txt)