mirror of
https://github.com/katanemo/plano.git
synced 2026-07-23 16:51:04 +02:00
feat(routing): decouple budget pricing from prompt_caching (#996)
* feat(routing): budget prices cache-aware regardless of prompt_caching flag * docs: prompt_caching is optional for the routing budget
This commit is contained in:
parent
66547bcef3
commit
ad92f7fbe7
8 changed files with 69 additions and 124 deletions
|
|
@ -547,8 +547,10 @@ properties:
|
|||
routing_budget:
|
||||
type: object
|
||||
description: >
|
||||
Per-session cost gate on model switching. Independent of prompt caching — it
|
||||
applies whenever configured (presence of this block turns it on). The default
|
||||
Per-session cost gate on model switching. Self-sufficient and independent of
|
||||
prompt_caching — it applies whenever configured (presence of this block turns
|
||||
it on), derives implicit sessions on its own, and always prices warm anchors
|
||||
at their cached input rate (provider caches are assumed real). The default
|
||||
posture is to stick to the model a session is warm on. When routing proposes a
|
||||
different model while that session's provider cache is plausibly still warm
|
||||
(inferred from the idle gap vs. the provider's cache window), the actual
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue