AgentDiet Context Compress — Web URL Distillation

Choose this endpoint when an AI agent needs to read or reference web content but must stay within a token budget, when you want to preprocess a URL before including it in a prompt, or when you need BM25-ranked passage extraction focused on a specific query. It is especially useful over generic scrapers when you care about downstream LLM cost savings and need structured ROI metrics on compression. Prefer it over raw HTTP fetch + manual truncation when you want mode-driven distillation (markdown, structured, skeleton, or budget) in a single call.

• URL is unreachable or returns a non-200 HTTP status — service may return an error status
• URL resolves to a page with no extractable text (e.g. fully JavaScript-rendered SPA with no SSR) — output may be empty or minimal
• max_tokens set too low (below 20) — request rejected by schema validation
• mode=budget used without providing a query — BM25 scoring may be unfocused
• Payment not authorized via x402 — 402 Payment Required response blocking the call
• Raw HTML provided is malformed or too large — may cause truncation or parse errors