sendnrw/netbird
2
0
Fork 0
mirror of https://github.com/netbirdio/netbird.git synced 2026-09-30 18:49:08 +02:00
Code Issues Packages Projects Releases Wiki Activity

[management] expose prompt-cache accounting in the agent-network REST API

Browse Source 
Access-log entries, session summaries, and usage-overview buckets gain
cached_input_tokens, cache_creation_tokens, and cache_cost_usd so the
cache share of a request's cost is visible instead of hiding inside
total_tokens. Generated with shared/management/http/api/generate.sh.
This commit is contained in:
Maycon Santos
2026-07-25 19:25:24 +00:00
parent d9bf667275
commit 0c26e364eb
2 changed files with 87 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
shared/management/http/api/openapi.yml
+57 -3
View File
@@ -5822,13 +5822,28 @@ components:
total_tokens:
type: integer
format: int64
description: Total tokens consumed.
description: Total tokens consumed, including prompt-cache tokens.
example: 1840
cached_input_tokens:
type: integer
format: int64
description: Input tokens read from the provider's prompt cache. Additive to input_tokens for Anthropic-shape providers; a subset of input_tokens for OpenAI.
example: 0
cache_creation_tokens:
type: integer
format: int64
description: Input tokens written to the provider's prompt cache. Zero for providers without a cache-write bucket.
example: 30528
cost_usd:
type: number
format: double
description: Estimated USD cost of the request.
example: 0.0231
cache_cost_usd:
type: number
format: double
description: Portion of cost_usd billed for prompt-cache usage.
example: 0.1145
stream:
type: boolean
description: Whether the request was a streaming completion.
@@ -5852,7 +5867,10 @@ components:
- input_tokens
- output_tokens
- total_tokens
- cached_input_tokens
- cache_creation_tokens
- cost_usd
- cache_cost_usd
AgentNetworkAccessLogsResponse:
type: object
properties:
@@ -5926,13 +5944,28 @@ components:
total_tokens:
type: integer
format: int64
description: Total tokens across the session.
description: Total tokens across the session, including prompt-cache tokens.
example: 12880
cached_input_tokens:
type: integer
format: int64
description: Total prompt-cache read tokens across the session.
example: 0
cache_creation_tokens:
type: integer
format: int64
description: Total prompt-cache write tokens across the session.
example: 30528
cost_usd:
type: number
format: double
description: Total estimated USD cost across the session.
example: 0.1617
cache_cost_usd:
type: number
format: double
description: Portion of cost_usd billed for prompt-cache usage across the session.
example: 0.1145
providers:
type: array
items:
@@ -5959,7 +5992,10 @@ components:
- input_tokens
- output_tokens
- total_tokens
- cached_input_tokens
- cache_creation_tokens
- cost_usd
- cache_cost_usd
- decision
- entries
AgentNetworkAccessLogSessionsResponse:
@@ -6013,19 +6049,37 @@ components:
total_tokens:
type: integer
format: int64
description: Total tokens in the bucket.
description: Total tokens in the bucket, including prompt-cache tokens.
example: 184000
cached_input_tokens:
type: integer
format: int64
description: Total prompt-cache read tokens in the bucket.
example: 20000
cache_creation_tokens:
type: integer
format: int64
description: Total prompt-cache write tokens in the bucket.
example: 45000
cost_usd:
type: number
format: double
description: Total estimated USD spend in the bucket.
example: 2.31
cache_cost_usd:
type: number
format: double
description: Portion of cost_usd billed for prompt-cache usage in the bucket.
example: 0.42
required:
- period_start
- input_tokens
- output_tokens
- total_tokens
- cached_input_tokens
- cache_creation_tokens
- cost_usd
- cache_cost_usd
AgentNetworkSettings:
type: object
description: Per-account Agent Network gateway settings. One row per account; cluster and subdomain are auto-assigned on first provider create and immutable thereafter.