Preset

Inbox copilot · document_extracted · version 3

Cache

Exact-match response cache

Semantic cache: off in this sample. Turn it on for routes where similar questions can share one answer.

Document extract follows a quality-first policy.
A served response has to score at least 0.86 and finish inside 2,200 ms at p95.
Anything priced above $0.0400 a request is outside this policy.
Simple requests can move from anthropic/claude-sonnet-4.5 to anthropic/claude-haiku-4.5 only after a shadow test passes.
Cascade is off, so there is no cheap-first attempt.
If the chosen model fails, fallbacks run in order: anthropic/claude-haiku-4.5, anthropic/claude-sonnet-4.5.
Output is capped at 1,200 tokens and must be structured, with reasoning effort low.
The route is synchronous, so it stays off batch prices.
Exact-match response cache is opt-in and on for this route.
The near-duplicate (semantic) cache runs only on routes you list as safe for it.
Parezen only suggests. A person still chooses Enable or Ignore.

This sample has no cascade escalations. When one happens, its cost is included, not hidden.
Exact-match cache is on. This sample has no exact-match hits yet, so the projection does not invent one.
The near-duplicate (semantic) cache is left out of this projection; it runs only on routes you list.