litellm/ui
Tin Chi Lo f7aecc3354 feat(proxy): estimate prompt-cache switch costs
Add a two-arm input-cost quote for an assumed cache allocation.
Each arm carries uncached, cache-read, five-minute write and
one-hour write token buckets, so an operator can compare staying
on a warm model with a cold switch using the configured deployment's
pricing. The endpoint rejects ambiguous model groups unless a concrete
deployment id identifies the arm and does not influence routing.

Co-Authored-By: Claude Code <noreply@anthropic.com>
2026-09-11 19:34:24 -07:00
..
litellm-dashboard feat(proxy): estimate prompt-cache switch costs 2026-09-11 19:34:24 -07:00
Dockerfile fix(docker): bump nginx runtime to 1.31.5-alpine3.24 and pin digest to resolve critical CVEs (#39561) 2026-09-03 08:46:07 -07:00
nginx.conf fix(ui): boot the UI image as an arbitrary uid by anchoring nginx writes under /tmp (#37982) 2026-08-24 11:57:36 -07:00