Benchmark report
Not supported
0m 24s elapsed
Official documentation shows this product does not offer the task. Parallel's documented web APIs (Search, Extract, Task/Deep Research, Responses, FindAll, Monitor, Ingest) include no multi-page crawl endpoint. The documentation index at docs.parallel.ai/llms.txt lists every product and no crawl API appears. Extract (/v1/extract) only fetches an explicitly supplied list of URLs; its full parameter set (urls, objective, search_queries, advanced_settings with fetch_policy, excerpt_settings, full_content) contains no crawl-start URL, no link-following/depth option, no page cap for discovered pages, and no URL path include/prefix/regex filter restricting returned URLs. The 'Crawler' page is webmaster guidance about Parallel's own crawler, not a customer-facing crawl API. Therefore a single product-side crawl job starting at https://www.postgresql.org/docs/17/tutorial.html with a documented path filter and a pre-set page cap of 25 cannot be run; doing it would require following links in client code, which the task forbids.
Where the time went
0:24 agent time- Starting0:01
- Research0:24
Attempt record
- —Not supported by this productParallel's documented web APIs (Search, Extract, Task/Deep Research, Responses, FindAll, Monitor, Ingest) include no multi-page crawl endpoint. The documentation index at docs.parallel.ai/llms.txt lists every product and no crawl API appears. Extract (/v1/extract) only fetches an explicitly supplied list of URLs; its full parameter set (urls, objective, search_queries, advanced_settings with fetch_policy, excerpt_settings, full_content) contains no crawl-start URL, no link-following/depth option, no page cap for discovered pages, and no URL path include/prefix/regex filter restricting returned URLs. The 'Crawler' page is webmaster guidance about Parallel's own crawler, not a customer-facing crawl API. Therefore a single product-side crawl job starting at https://www.postgresql.org/docs/17/tutorial.html with a documented path filter and a pre-set page cap of 25 cannot be run; doing it would require following links in client code, which the task forbids.Parallel documentation index (llms.txt)Extract API QuickstartAdvanced Extract Settings