AgentDiet Web Distiller — Compact Token Extraction from URLs

Choose this endpoint when you need to feed web content to an LLM but want to minimize token consumption and cost. It is especially valuable when the target page is long and you only need a focused excerpt (use BM25 budget mode with a query), or when you want structured JSON tables/metadata rather than raw HTML. Prefer it over generic scraping tools when token compression ratios and LLM cost savings are a priority, or when you need the output in a directly agent-consumable format (markdown, skeleton outline, or structured JSON).

• URL is unreachable or returns a non-200 HTTP response — endpoint may return an error status
• HTML too large or malformed — distillation may truncate or fail
• mode=budget without a query or max_tokens — may default to full markdown without focusing
• Payment (x402) not completed — returns HTTP 402 requiring USDC payment before proceeding
• Invalid URL format — schema requires minLength 1 but malformed URLs may cause fetch errors
• Rate limiting or timeout on slow external pages — latency spike or timeout error