Xch13 6ba20acd66 fix(frontend): correct Cache Hit Rate denominator to include all prompt tokens
Cache Hit Rate was calculated as cache_read / (cache_read + cache_creation),
which always yields 100% for OpenAI models since cache_creation is never
reported by the OpenAI API. The denominator should include all prompt tokens
(input_tokens + cache_read_tokens + cache_creation_tokens) so the rate
reflects the actual percentage of input tokens served from cache.

Fixes #2291
2026-05-22 11:45:33 +08:00
..
2025-12-18 14:26:55 +08:00