mirror of
https://github.com/BerriAI/litellm.git
synced 2026-10-07 02:59:05 +00:00
Update test_llm_cost_calc_utils.py
This commit is contained in:
parent
0bcefacdf5
commit
34cf1f5bac
1 changed files with 0 additions and 38 deletions
|
|
@ -172,44 +172,6 @@ def test_inline_pdf_tokens_costed_when_text_tokens_less_than_prompt_tokens():
|
|||
f"got cost={cost}. PDF tokens are likely not being costed."
|
||||
)
|
||||
|
||||
def test_inline_pdf_tokens_costed_when_text_both_costed_correctly():
|
||||
usage = Usage(
|
||||
completion_tokens=80,
|
||||
prompt_tokens=6000,
|
||||
total_tokens=6080,
|
||||
prompt_tokens_details=PromptTokensDetailsWrapper(
|
||||
audio_tokens=500, cached_tokens=None, text_tokens=50, image_tokens=None
|
||||
),
|
||||
)
|
||||
response = ModelResponse(
|
||||
usage= usage,
|
||||
model = "gemini-2.0-flash-001",
|
||||
)
|
||||
cost = response_cost_calculator(
|
||||
response_object= response,
|
||||
model = "gemini-2.0-flash-001",
|
||||
custom_llm_provider = "vertex_ai",
|
||||
call_type = "acompletion",
|
||||
optional_params= {},
|
||||
)
|
||||
|
||||
model_info = litellm.model_cost["gemini-2.0-flash-001"]
|
||||
input_cost_per_token = model_info["input_cost_per_token"]
|
||||
input_cost_per_audio_token = model_info["input_cost_per_audio_token"]
|
||||
output_cost_per_token = model_info["output_cost_per_token"]
|
||||
|
||||
expected_input_cost = (
|
||||
5500 * input_cost_per_token + 500 * input_cost_per_audio_token
|
||||
)
|
||||
expected_output_cost = 80 * output_cost_per_token
|
||||
expected_cost = expected_input_cost + expected_output_cost
|
||||
|
||||
assert cost == pytest.approx(expected_cost), (
|
||||
f"Expected cost={expected_cost}, got cost={cost}. "
|
||||
f"Unaccounted PDF tokens alongside audio are not being costed."
|
||||
)
|
||||
|
||||
|
||||
def test_inline_pdf_with_audio_tokens_both_costed_correctly():
|
||||
"""
|
||||
Scenario: User sends a PDF inline + audio + text.
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue