scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-12 23:01:41 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 182

fix: apply cache_read_input_token_cost to cached tokens in Fireworks AI cost calculation

Browse source
This commit is contained in:
PRABHU KIRAN VANDRANKI 2026-04-18 13:21:31 -04:00
parent 850fe595ac
commit d0b62e79fb
1 changed files with 23 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

24
litellm/llms/fireworks_ai/cost_calculator.py
View file

@ -77,10 +77,32 @@ def cost_per_token(model: str, usage: Usage) -> Tuple[float, float]:
)
## CALCULATE INPUT COST
# Extract cached and cache-creation tokens from usage details
cached_tokens = 0
cache_creation_tokens = 0
if getattr(usage, "prompt_tokens_details", None) is not None:
cached_tokens = (
getattr(usage.prompt_tokens_details, "cached_tokens", 0) or 0
)
cache_creation_tokens = (
getattr(usage.prompt_tokens_details, "cache_creation_tokens", 0) or 0
)
prompt_cost: float = usage["prompt_tokens"] * model_info["input_cost_per_token"]
cache_read_cost: float = model_info.get("cache_read_input_token_cost") or 0.0
cache_creation_cost: float = (
model_info.get("cache_creation_input_token_cost") or 0.0
)
# Non-cached tokens are billed at the standard input rate
non_cached_tokens = usage["prompt_tokens"] - cached_tokens - cache_creation_tokens
prompt_cost: float = (
non_cached_tokens * model_info["input_cost_per_token"]
+ cached_tokens * cache_read_cost
+ cache_creation_tokens * cache_creation_cost
)
## CALCULATE OUTPUT COST
completion_cost = usage["completion_tokens"] * model_info["output_cost_per_token"]
return prompt_cost, completion_cost