Web URL Context Compressor (BM25 Budget Control)

Choose this endpoint when you need to feed web page content into an LLM context window but want to minimize token usage and cost. It is especially valuable when you have a strict token budget (use mode=budget with a BM25 query to surface only the most relevant passage), when you need structured tables or metadata from a page (mode=structured), or when a high-level outline suffices (mode=skeleton). Prefer it over raw web scraping when downstream LLM token cost matters, and when you want quantified proof of compression savings (ROI multiplier and cost saved are returned with every response).

• URL is unreachable or returns non-200 HTTP status — endpoint may return an error status
• URL points to non-HTML content (PDF, binary) — extraction may fail or return empty data
• Token budget too low (max_tokens below 20) — rejected by schema validation
• BM25 query provided without mode=budget — query parameter is silently ignored
• Raw HTML and URL both omitted — URL is required and missing will cause a validation error
• Network timeout fetching the target URL — may return a timeout error
• Payment not completed via x402 — returns HTTP 402 with payment instructions