Implement prompt caching, speculative decoding config, parallel slots, and dynamic pricing metrics

This commit is contained in:
Hitonabi
2026-06-26 14:00:32 +02:00
parent 2c60caf790
commit 35dcc69ba5
12 changed files with 216 additions and 35 deletions
+3
View File
@@ -86,6 +86,9 @@ export interface ModelInfo {
quant: string
size_bytes: number | null
incomplete: boolean
prompt_cache: boolean
spec_draft_model: string | null
parallel_slots: number
capabilities: Capabilities
}