scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-11 22:51:28 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 154

test(cost): cover reasoning nested in text_tokens beside audio output

Browse source
This commit is contained in:
mateo-berri 2026-09-04 20:35:21 -07:00
parent 5a7e938f37
commit 92e449d4d0
1 changed files with 19 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

19
tests/test_litellm/litellm_core_utils/llm_cost_calc/test_llm_cost_calc_utils.py
View file

@ -4774,3 +4774,22 @@ def test_generic_cost_per_token_keeps_billing_reasoning_reported_beside_text_tok
info = litellm.get_model_info(model=model, custom_llm_provider="openai")
assert completion_cost == pytest.approx(44 * info["output_cost_per_token"])
def test_generic_cost_per_token_bills_nested_reasoning_once_beside_audio_output(_local_model_cost_map: None) -> None:
"""Audio-output realtime usage nests reasoning inside text_tokens next to audio_tokens; text is billed net of it."""
model = "gpt-realtime-2.1-mini"
usage = Usage(
prompt_tokens=120,
completion_tokens=100,
total_tokens=220,
completion_tokens_details=CompletionTokensDetailsWrapper(text_tokens=30, audio_tokens=70, reasoning_tokens=20),
)
_, completion_cost = generic_cost_per_token(model=model, usage=usage, custom_llm_provider="openai")
breakdown = get_token_type_cost_breakdown(model=model, custom_llm_provider="openai", usage=usage)
info = litellm.get_model_info(model=model, custom_llm_provider="openai")
assert breakdown.reasoning_cost == pytest.approx(20 * info["output_cost_per_token"])
assert completion_cost == pytest.approx(30 * info["output_cost_per_token"] + 70 * info["output_cost_per_audio_token"])