scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-11-11 08:38:24 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 280

fix(cost): keep the token breakdown when a header reported cost repeats an already priced bill

Browse source
This commit is contained in:
devin-ai-integration[bot] 2026-10-29 20:44:32 +00:00
parent b324599f0d
commit 9a1e0f868d
2 changed files with 17 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

23
litellm/cost_calculator.py
View file

@ -1505,6 +1505,12 @@ def _response_time_ms_for_cost(
return 0.0
def _breakdown_total(litellm_logging_obj: LitellmLoggingObject | None) -> float | None:
breakdown: Final[object] = getattr(litellm_logging_obj, "cost_breakdown", None)
total: Final[object] = breakdown.get("total_cost") if isinstance(breakdown, Mapping) else None
return float(total) if isinstance(total, (int, float)) and not isinstance(total, bool) else None
def _adjusted_provider_billed_cost(
reported_cost: float,
custom_llm_provider: str | None,
@ -2334,14 +2340,15 @@ def response_cost_calculator(
)
provider_response_cost: Final = get_response_cost_from_hidden_params(hidden_params)
if provider_response_cost is not None and not custom_pricing:
_store_cost_breakdown_in_logging_obj(
litellm_logging_obj=litellm_logging_obj,
prompt_tokens_cost_usd_dollar=0.0,
completion_tokens_cost_usd_dollar=provider_response_cost,
cost_for_built_in_tools_cost_usd_dollar=0.0,
total_cost_usd_dollar=provider_response_cost,
original_cost=provider_response_cost,
)
if _breakdown_total(litellm_logging_obj) != provider_response_cost:
_store_cost_breakdown_in_logging_obj(
litellm_logging_obj=litellm_logging_obj,
prompt_tokens_cost_usd_dollar=0.0,
completion_tokens_cost_usd_dollar=provider_response_cost,
cost_for_built_in_tools_cost_usd_dollar=0.0,
total_cost_usd_dollar=provider_response_cost,
original_cost=provider_response_cost,
)
return provider_response_cost
response_cost = completion_cost(

3
litellm/litellm_core_utils/core_helpers.py
View file

@ -1028,7 +1028,8 @@ def mark_usage_cost_as_litellm_bill(response: object, bill: float) -> None:
so ``billed_usage_cost`` does not read it back as the upstream's charge on a later pricing pass."""
hidden: Final[object] = getattr(response, "_hidden_params", None)
if isinstance(hidden, dict):
cast(dict[str, object], hidden).update({"response_cost": bill, LITELLM_BILLED_USAGE_COST: bill})
hidden_params: Final = cast(dict[str, object], hidden) # cast-ok: hidden params ride on the response as a dict
hidden_params.update({"response_cost": bill, LITELLM_BILLED_USAGE_COST: bill})
def _usage_cost_is_litellm_bill(response: object) -> bool: