scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-10-10 03:28:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 85

fix: only recalculate total_tokens when cache tokens present

Browse source
This commit is contained in:
Monesh Ram 2026-02-24 09:37:38 +05:30 • committed by GitHub
parent fe9a6458a5
commit 4a024957ec
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 4 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
litellm/llms/bedrock/chat/converse_transformation.py
View file

@ -1644,9 +1644,10 @@ class AmazonConverseConfig(BaseConfig):
if "cacheWriteInputTokens" in usage:
cache_creation_input_tokens = usage["cacheWriteInputTokens"]
input_tokens += cache_creation_input_tokens
# Recalculate total_tokens to include cache tokens added to input_tokens
# The API's totalTokens doesn't include cache hits/writes, but prompt_tokens does
total_tokens = input_tokens + output_tokens
# Only recalculate total_tokens if we added cache tokens to input_tokens
# Otherwise, trust the API's totalTokens (which may include other token categories)
if cache_read_input_tokens > 0 or cache_creation_input_tokens > 0:
total_tokens = input_tokens + output_tokens
prompt_tokens_details = PromptTokensDetailsWrapper(
cached_tokens=cache_read_input_tokens
)