scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-10-26 06:03:19 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 35
litellm/tests/integration/cost_calculation/cases/misc.py
devin-ai-integration[bot] cd20ecace1
test(cost): add tier 1 cost tracking cells and sync usage shapes for 2026-10-17 (#47141) 
* test(cost): add tier 1 cost tracking cells and sync usage shapes for 2026-10-17

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* test(cost): restore tiered cost cases and revert out of scope usage edits

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* test(cost): add remaining tier 1 endpoint cells for 2026-10-17

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

---------

Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-10-17 11:11:46 -07:00

2218 lines
85 KiB
Python
Raw Blame History

"""azure_ai, bedrock, cohere, cohere_chat, dashscope, deepgram, deepseek, groq, mistral, openrouter, text-completion-openai, xai cost tracking cases, one CostTrackingTestCase literal per request shape (moved from cost_tracking_cases.json).
PARITY pairs the streamed case with its plain twin where both bill the same row."""
from typing import Final, cast
from pydantic import JsonValue
from integration.cost_calculation.cost_tracking_case import (
ALWAYS_OPEN_OFF_PEAK_WINDOW,
OffPeakPricing,
CostTrackingTestCase,
Deployment,
EventStreamEvent,
EventStreamResponse,
ExactExpected,
JsonResponse,
RecountExpected,
RecountRates,
SseResponse,
WavUpload,
with_usage,
)
from integration.cost_calculation.stream_parity.case import StreamParityTestCase, sse_frames
from integration.cost_calculation.bridge_cost_parity.case import BridgeCostParityTestCase, bridge_pairs
from integration.cost_calculation.cases.anthropic import (
CLAUDE_FABLE_5_1_CACHE_READ,
CLAUDE_FABLE_5_1_CACHE_WRITE_1H,
CLAUDE_FABLE_5_1_CACHE_WRITE_5M,
CLAUDE_FABLE_5_1_INPUT_TEXT,
CLAUDE_FABLE_5_1_MESSAGES_INPUT_TEXT,
CLAUDE_FABLE_5_1_MESSAGES_STREAM,
CLAUDE_FABLE_5_1_STREAM,
CLAUDE_FABLE_5_1_THINKING,
CLAUDE_FABLE_5_1_TOOL_CALL,
CLAUDE_HAIKU_5_5_CACHE_READ,
CLAUDE_HAIKU_5_5_CACHE_WRITE_1H,
CLAUDE_HAIKU_5_5_CACHE_WRITE_5M,
CLAUDE_HAIKU_5_5_INPUT_TEXT,
CLAUDE_HAIKU_5_5_MESSAGES_INPUT_TEXT,
CLAUDE_HAIKU_5_5_MESSAGES_STREAM,
CLAUDE_HAIKU_5_5_STREAM,
CLAUDE_HAIKU_5_5_THINKING,
CLAUDE_HAIKU_5_5_TOOL_CALL,
)
def _azure_ai_claude_messages_case(
response_case: CostTrackingTestCase,
request_case: CostTrackingTestCase,
name: str,
model: str,
) -> CostTrackingTestCase:
return response_case.model_copy(
update={
"name": name,
"model": model,
"endpoint": "/v1/messages",
"request": request_case.request,
}
)
def _azure_ai_claude_usage(
input_tokens: int,
output_tokens: int,
inference_geo: str,
*,
cache_creation_input_tokens: int = 0,
cache_read_input_tokens: int = 0,
ephemeral_5m_input_tokens: int = 0,
ephemeral_1h_input_tokens: int = 0,
thinking_tokens: int = 0,
) -> dict[str, JsonValue]:
return {
"input_tokens": input_tokens,
"cache_creation_input_tokens": cache_creation_input_tokens,
"cache_read_input_tokens": cache_read_input_tokens,
"cache_creation": {
"ephemeral_5m_input_tokens": ephemeral_5m_input_tokens,
"ephemeral_1h_input_tokens": ephemeral_1h_input_tokens,
},
"output_tokens": output_tokens,
"output_tokens_details": {"thinking_tokens": thinking_tokens},
"service_tier": "standard",
"inference_geo": inference_geo,
"speed": "standard",
}
def _azure_ai_claude_shape_case(
sibling: CostTrackingTestCase,
name: str,
model: str,
usage: dict[str, JsonValue],
expected: ExactExpected,
response_content: list[dict[str, JsonValue]] | None = None,
adaptive_thinking: bool = False,
) -> CostTrackingTestCase:
response: Final = cast(JsonResponse, sibling.response)
request: Final = (
{
**{
key: value
for key, value in sibling.request.items()
if key not in {"reasoning_effort", "allowed_openai_params"}
},
"thinking": {"type": "adaptive"},
"output_config": {"effort": "medium"},
"allowed_openai_params": ["thinking", "output_config"],
}
if adaptive_thinking
else sibling.request
)
body: Final = {
**response.body,
"model": model.rsplit("/", 1)[-1],
"usage": usage,
**({"content": response_content} if response_content is not None else {}),
}
return sibling.model_copy(
update={
"name": name,
"model": model,
"request": request,
"response": response.model_copy(update={"body": body}),
"expected": expected,
}
)
# nova-next
NOVA_NEXT_TRANSCRIPTIONS_PER_SECOND: Final = CostTrackingTestCase(
name="nova-next-transcriptions-per-second",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="nova-next",
endpoint="/v1/audio/transcriptions",
upload=WavUpload(kind="wav", seconds=4.0),
request={},
response=JsonResponse(
content_type="application/json",
body={
"results": {"channels": [{"alternatives": [{"transcript": "hello", "confidence": 0.9}]}]},
"metadata": {"duration": 4.0, "channels": 1},
},
),
expected=ExactExpected(spend=0.0012, input_cost=0.0012, output_cost=0, prompt_tokens=0, completion_tokens=0),
)
# amazon.nova-canvas-next
AMAZON_NOVA_CANVAS_NEXT_IMAGES_ONE: Final = CostTrackingTestCase(
name="amazon-nova-canvas-next-images-one",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="amazon.nova-canvas-next",
endpoint="/v1/images/generations",
deployment=Deployment(model="amazon.nova-canvas-next"),
request={"prompt": "a deterministic bedrock image"},
response=JsonResponse(content_type="application/json", body={"images": ["AA=="]}),
expected=ExactExpected(
spend=0.045, input_cost=0.045, output_cost=0, prompt_tokens=0, completion_tokens=0, breakdown_persisted=False
),
)
# embed-v5
COHERE_EMBEDDINGS_V5: Final = CostTrackingTestCase(
name="cohere-embeddings-v5",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="embed-v5",
endpoint="/v1/embeddings",
request={"model": "$MODEL", "input": "cohere embedding", "input_type": "search_query"},
response=JsonResponse(
content_type="application/json",
body={
"id": "emb-1",
"embeddings": {"float": [[0.1, 0.2, 0.3]]},
"meta": {"billed_units": {"input_tokens": 11}},
},
),
expected=ExactExpected(
spend=1.144e-05, input_cost=1.144e-05, output_cost=0.0, prompt_tokens=11, completion_tokens=0
),
)
# amazon.titan-embed-text-v2:0
BEDROCK_EMBEDDINGS_TITAN_V2: Final = CostTrackingTestCase(
name="bedrock-embeddings-titan-v2",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="amazon.titan-embed-text-v2:0",
endpoint="/v1/embeddings",
request={"model": "$MODEL", "input": "titan embedding"},
response=JsonResponse(
content_type="application/json", body={"embedding": [0.1, 0.2, 0.3], "inputTextTokenCount": 10}
),
expected=ExactExpected(spend=1.05e-05, input_cost=1.05e-05, output_cost=0.0, prompt_tokens=10, completion_tokens=0),
)
# cohere.embed-english-v4
BEDROCK_COHERE_EMBEDDINGS_V4: Final = CostTrackingTestCase(
name="bedrock-cohere-embeddings-v4",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="cohere.embed-english-v4",
endpoint="/v1/embeddings",
request={"model": "$MODEL", "input": "bedrock cohere embedding"},
response=JsonResponse(
content_type="application/json",
body={
"embeddings": [[0.1, 0.2, 0.3]],
"id": "emb-bedrock-cohere-1",
"response_type": "embeddings_floats",
"texts": ["bedrock cohere embedding"],
},
),
expected=ExactExpected(spend=5.3e-06, input_cost=5.3e-06, output_cost=0.0, prompt_tokens=5, completion_tokens=0),
)
# rerank-v4
COHERE_RERANK_V4_ONE: Final = CostTrackingTestCase(
name="cohere-rerank-v4-one",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="rerank-v4",
endpoint="/v1/rerank",
request={"model": "$MODEL", "query": "rank this", "documents": ["a", "b"]},
response=JsonResponse(
content_type="application/json",
body={
"id": "rr-$REQUEST_ID",
"results": [{"index": 0, "relevance_score": 0.9}],
"meta": {"api_version": {"version": "2"}, "billed_units": {"search_units": 1}},
},
),
expected=ExactExpected(spend=0.0021, input_cost=0.0021, output_cost=0.0, prompt_tokens=0, completion_tokens=0),
)
COHERE_RERANK_V4_THREE: Final = CostTrackingTestCase(
name="cohere-rerank-v4-three",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="rerank-v4",
endpoint="/v1/rerank",
request={
"model": "$MODEL",
"query": "rank this",
"documents": ["a long document", "another long document", "third long document"],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "rr-three-$REQUEST_ID",
"results": [{"index": 0, "relevance_score": 0.9}],
"meta": {"api_version": {"version": "2"}, "billed_units": {"search_units": 3}},
},
),
expected=ExactExpected(spend=0.0063, input_cost=0.0063, output_cost=0.0, prompt_tokens=0, completion_tokens=0),
)
COHERE_RERANK_V4_TOTAL_TOKENS_FALLBACK: Final = CostTrackingTestCase(
name="cohere-rerank-v4-total-tokens-fallback",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="rerank-v4",
endpoint="/v1/rerank",
request={"model": "$MODEL", "query": "rank this", "documents": ["fallback a", "fallback b"]},
response=JsonResponse(
content_type="application/json",
body={
"id": "rr-fallback-$REQUEST_ID",
"results": [{"index": 0, "relevance_score": 0.8}],
"meta": {"billed_units": {"total_tokens": 99}},
},
),
expected=ExactExpected(spend=0.0, input_cost=0.0, output_cost=0.0, prompt_tokens=0, completion_tokens=0),
)
# cohere.rerank-v4:0
BEDROCK_COHERE_RERANK_V4: Final = CostTrackingTestCase(
name="bedrock-cohere-rerank-v4",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="cohere.rerank-v4:0",
endpoint="/v1/rerank",
request={"model": "$MODEL", "query": "rank this", "documents": ["a", "b"]},
response=JsonResponse(
content_type="application/json",
body={"results": [{"index": 0, "relevanceScore": 0.9}], "response_id": "rr-3", "token_count": 1},
),
expected=ExactExpected(spend=0.0022, input_cost=0.0022, output_cost=0.0, prompt_tokens=0, completion_tokens=0),
)
# gpt-3.5-turbo-instruct-next
TEXT_COMPLETIONS_OPENAI_BASIC: Final = CostTrackingTestCase(
name="text-completions-openai-basic",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="gpt-3.5-turbo-instruct-next",
endpoint="/v1/completions",
request={"model": "$MODEL", "prompt": "complete this"},
response=JsonResponse(
content_type="application/json",
body={
"id": "cmpl-basic-$REQUEST_ID",
"object": "text_completion",
"choices": [{"text": "done", "index": 0, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 9, "completion_tokens": 4, "total_tokens": 13},
},
),
expected=ExactExpected(
spend=1.882e-05, input_cost=1.026e-05, output_cost=8.56e-06, prompt_tokens=9, completion_tokens=4
),
)
TEXT_COMPLETIONS_OPENAI_STREAM_USAGE: Final = CostTrackingTestCase(
name="text-completions-openai-stream-usage",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="gpt-3.5-turbo-instruct-next",
endpoint="/v1/completions",
request={"model": "$MODEL", "prompt": "complete this", "stream": True, "stream_options": {"include_usage": True}},
response=SseResponse(
content_type="text/event-stream",
frames=sse_frames(
{
"id": "cmpl-$REQUEST_ID",
"object": "text_completion",
"created": 1789789000,
"model": "gpt-3.5-turbo-instruct-next",
"choices": [{"text": "done", "index": 0, "finish_reason": None}],
"usage": None,
},
{
"id": "cmpl-$REQUEST_ID",
"object": "text_completion",
"created": 1789789000,
"model": "gpt-3.5-turbo-instruct-next",
"choices": [],
"usage": {"prompt_tokens": 9, "completion_tokens": 4, "total_tokens": 13},
},
done=True,
),
),
expected=ExactExpected(
spend=1.882e-05, input_cost=1.026e-05, output_cost=8.56e-06, prompt_tokens=9, completion_tokens=4
),
)
TEXT_COMPLETIONS_OPENAI_N_BEST: Final = CostTrackingTestCase(
name="text-completions-openai-n-best",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="gpt-3.5-turbo-instruct-next",
endpoint="/v1/completions",
request={"model": "$MODEL", "prompt": "complete this twice", "n": 2},
response=JsonResponse(
content_type="application/json",
body={
"id": "cmpl-n-best-$REQUEST_ID",
"object": "text_completion",
"choices": [
{"text": "done", "index": 0, "finish_reason": "stop"},
{"text": "also done", "index": 1, "finish_reason": "stop"},
],
"usage": {"prompt_tokens": 9, "completion_tokens": 8, "total_tokens": 17},
},
),
expected=ExactExpected(
spend=2.738e-05, input_cost=1.026e-05, output_cost=1.712e-05, prompt_tokens=9, completion_tokens=8
),
)
# dashscope/qwen4-max
DASHSCOPE_QWEN4_MAX_TIERED_INPUT: Final = CostTrackingTestCase(
name="dashscope-qwen4-max-tiered_input",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="dashscope/qwen4-max",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "tiered input"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "$REQUEST_ID",
"object": "chat.completion",
"model": "qwen4-max",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
},
),
expected=ExactExpected(
spend=0.00507, input_cost=0.002392, output_cost=0.002678, prompt_tokens=1840, completion_tokens=412
),
)
DASHSCOPE_QWEN4_MAX_TIERED_BOUNDARY_STAYS_LOWER_TIER: Final = CostTrackingTestCase(
name="dashscope-qwen4-max-tiered_boundary_stays_lower_tier",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="dashscope/qwen4-max",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "tier boundary"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "$REQUEST_ID",
"object": "chat.completion",
"model": "qwen4-max",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 32000, "completion_tokens": 412, "total_tokens": 32412},
},
),
expected=ExactExpected(
spend=0.044278, input_cost=0.0416, output_cost=0.002678, prompt_tokens=32000, completion_tokens=412
),
)
DASHSCOPE_QWEN4_MAX_TIERED_SECOND_TIER: Final = CostTrackingTestCase(
name="dashscope-qwen4-max-tiered_second_tier",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="dashscope/qwen4-max",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "tier two"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "$REQUEST_ID",
"object": "chat.completion",
"model": "qwen4-max",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 40000, "completion_tokens": 412, "total_tokens": 40412},
},
),
expected=ExactExpected(
spend=0.109356, input_cost=0.104, output_cost=0.005356, prompt_tokens=40000, completion_tokens=412
),
)
DASHSCOPE_QWEN4_MAX_TIERED_ABOVE_TOP_RANGE: Final = CostTrackingTestCase(
name="dashscope-qwen4-max-tiered_above_top_range",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="dashscope/qwen4-max",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "top tier"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "$REQUEST_ID",
"object": "chat.completion",
"model": "qwen4-max",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 300000, "completion_tokens": 412, "total_tokens": 300412},
},
),
expected=ExactExpected(
spend=0.936386, input_cost=0.93, output_cost=0.006386, prompt_tokens=300000, completion_tokens=412
),
)
# openrouter/anthropic/claude-sonnet-5
OPENROUTER_ANTHROPIC_CLAUDE_SONNET_5_PROVIDER_REPORTED_COST: Final = CostTrackingTestCase(
name="openrouter-anthropic-claude-sonnet-5-provider_reported_cost",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="openrouter/anthropic/claude-sonnet-5",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "reported cost"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "anthropic/claude-sonnet-5",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252, "cost": 0.0421},
},
),
expected=ExactExpected(
spend=0.0421,
input_cost=0.0,
output_cost=0.0421,
prompt_tokens=1840,
completion_tokens=412,
breakdown_persisted=False,
),
)
OPENROUTER_ANTHROPIC_CLAUDE_SONNET_5_TOKEN_PRICED: Final = CostTrackingTestCase(
name="openrouter-anthropic-claude-sonnet-5-token_priced",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="openrouter/anthropic/claude-sonnet-5",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "token pricing"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "anthropic/claude-sonnet-5",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
},
),
expected=ExactExpected(
spend=0.01248, input_cost=0.005888, output_cost=0.006592, prompt_tokens=1840, completion_tokens=412
),
)
# perplexity/sonar-next
PERPLEXITY_SONAR_NEXT_NO_SEARCH: Final = CostTrackingTestCase(
name="perplexity-sonar-next-no_search",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="perplexity/sonar-next",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "no search"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "sonar-next",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
},
),
expected=ExactExpected(
spend=0.0025118, input_cost=0.0020792, output_cost=0.0004326, prompt_tokens=1840, completion_tokens=412
),
)
# deepseek/deepseek-v4-chat
DEEPSEEK_DEEPSEEK_V4_CHAT_PROMPT_CACHE_HIT: Final = CostTrackingTestCase(
name="deepseek-deepseek-v4-chat-prompt_cache_hit",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="deepseek/deepseek-v4-chat",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "cache hit"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "deepseek-v4-chat",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {
"prompt_tokens": 1840,
"completion_tokens": 412,
"total_tokens": 2252,
"prompt_cache_hit_tokens": 1200,
"prompt_cache_miss_tokens": 640,
"prompt_tokens_details": {"cached_tokens": 1200},
},
},
),
expected=ExactExpected(
spend=0.00039756,
input_cost=0.0002204,
output_cost=0.00017716,
cache_read_cost=3.48e-05,
prompt_tokens=1840,
completion_tokens=412,
),
)
DEEPSEEK_DEEPSEEK_V4_CHAT_NO_CACHE_FIELDS_BILLS_ZERO_CACHE: Final = CostTrackingTestCase(
name="deepseek-deepseek-v4-chat-no_cache_fields_bills_zero_cache",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="deepseek/deepseek-v4-chat",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "no cache"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "deepseek-v4-chat",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
},
),
expected=ExactExpected(
spend=0.00071076,
input_cost=0.0005336,
output_cost=0.00017716,
cache_read_cost=0.0,
prompt_tokens=1840,
completion_tokens=412,
),
)
# xai/grok-5
XAI_GROK_5_REASONING_FOLDED_INTO_COMPLETION: Final = CostTrackingTestCase(
name="xai-grok-5-reasoning_folded_into_completion",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="xai/grok-5",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "reasoning"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "grok-5",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {
"prompt_tokens": 1840,
"completion_tokens": 412,
"total_tokens": 2552,
"completion_tokens_details": {"reasoning_tokens": 300},
},
},
),
expected=ExactExpected(
spend=0.0044064, input_cost=0.002484, output_cost=0.0019224, prompt_tokens=1840, completion_tokens=712
),
)
XAI_GROK_5_LIVE_SEARCH: Final = CostTrackingTestCase(
name="xai-grok-5-live_search",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="xai/grok-5",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "live search"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "grok-5",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {
"prompt_tokens": 1840,
"completion_tokens": 412,
"total_tokens": 2252,
"server_side_tool_usage_details": {"web_search_calls": 2},
},
},
),
expected=ExactExpected(
spend=0.0135964,
input_cost=0.002484,
output_cost=0.0011124,
tool_usage_cost=0.01,
prompt_tokens=1840,
completion_tokens=412,
),
)
XAI_GROK_5_PROVIDER_REPORTED_COST: Final = CostTrackingTestCase(
name="xai-grok-5-provider_reported_cost",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="xai/grok-5",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "reported xai cost"}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "grok-5",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252, "cost": 0.0421},
},
),
expected=ExactExpected(
spend=0.0421,
input_cost=0.029078078,
output_cost=0.013021922,
prompt_tokens=1840,
completion_tokens=412,
),
)
# bedrock/invoke/anthropic.claude-haiku-4-5-20251001-v1:0
BEDROCK_INVOKE_HAIKU_JSON: Final = CostTrackingTestCase(
name="bedrock-invoke-haiku-json",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="bedrock/invoke/anthropic.claude-haiku-4-5-20251001-v1:0",
endpoint="/v1/messages",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "bedrock-invoke-haiku-json"}],
"stream": False,
"max_tokens": 412,
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "anthropic.claude-haiku-4-5-20251001-v1:0",
"content": [{"type": "text", "text": "scripted response"}],
"stop_reason": "end_turn",
"stop_sequence": None,
"usage": {"input_tokens": 1840, "output_tokens": 412},
},
),
expected=ExactExpected(
spend=0.00425372, input_cost=0.0021896, output_cost=0.00206412, prompt_tokens=1840, completion_tokens=412
),
)
BEDROCK_INVOKE_HAIKU_STREAM: Final = CostTrackingTestCase(
name="bedrock-invoke-haiku-stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="bedrock/invoke/anthropic.claude-haiku-4-5-20251001-v1:0",
endpoint="/v1/messages",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "bedrock-invoke-haiku-stream"}],
"stream": True,
"max_tokens": 412,
},
response=EventStreamResponse(
content_type="application/vnd.amazon.eventstream",
events=(
EventStreamEvent(
event_type="message_start",
payload={
"type": "message_start",
"message": {
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "anthropic.claude-haiku-4-5-20251001-v1:0",
"content": [],
"stop_reason": None,
"stop_sequence": None,
"usage": {"input_tokens": 1840, "output_tokens": 0},
},
},
),
EventStreamEvent(
event_type="content_block_delta",
payload={
"type": "content_block_delta",
"index": 0,
"delta": {"type": "text_delta", "text": "scripted response"},
},
),
EventStreamEvent(
event_type="message_delta",
payload={
"type": "message_delta",
"delta": {"stop_reason": "end_turn"},
"usage": {"output_tokens": 412},
},
),
EventStreamEvent(event_type="message_stop", payload={"type": "message_stop"}),
),
framing="invoke",
),
expected=ExactExpected(
spend=0.00425372, input_cost=0.0021896, output_cost=0.00206412, prompt_tokens=1840, completion_tokens=412
),
)
BEDROCK_INVOKE_HAIKU_CHAT_STREAM: Final = CostTrackingTestCase(
name="bedrock-invoke-haiku-chat-stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="bedrock/invoke/anthropic.claude-haiku-4-5-20251001-v1:0",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "bedrock-invoke-haiku-chat-stream"}],
"stream": True,
"stream_options": {"include_usage": True},
"max_tokens": 412,
},
response=BEDROCK_INVOKE_HAIKU_STREAM.response,
expected=BEDROCK_INVOKE_HAIKU_STREAM.expected,
)
BEDROCK_INVOKE_HAIKU_RESPONSES: Final = CostTrackingTestCase(
name="bedrock-invoke-haiku-responses",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="bedrock/invoke/anthropic.claude-haiku-4-5-20251001-v1:0",
endpoint="/v1/responses",
request={"model": "$MODEL", "input": "bedrock-invoke-haiku-responses", "max_output_tokens": 412},
response=BEDROCK_INVOKE_HAIKU_JSON.response,
expected=BEDROCK_INVOKE_HAIKU_JSON.expected,
)
# azure_ai/gpt-5.4-mini-2026-03-17
AZURE_AI_GPT_5_4_MINI_LATEST: Final = CostTrackingTestCase(
name="azure-ai-gpt-5.4-mini-latest",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/gpt-5.4-mini-2026-03-17",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "azure-ai-gpt-5.4-mini-latest"}],
"stream": False,
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "$MODEL",
"choices": [
{"index": 0, "message": {"role": "assistant", "content": "scripted response"}, "finish_reason": "stop"}
],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
"service_tier": "default",
},
),
expected=ExactExpected(
spend=0.003234, input_cost=0.00138, output_cost=0.001854, prompt_tokens=1840, completion_tokens=412
),
)
AZURE_AI_GPT_5_4_MINI_LATEST_STREAM: Final = CostTrackingTestCase(
name="azure-ai-gpt-5.4-mini-latest-stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="azure_ai/gpt-5.4-mini-2026-03-17",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "azure-ai-gpt-5.4-mini-latest-stream"}],
"stream": True,
},
response=SseResponse(
content_type="text/event-stream",
frames=(
'data: {"id":"chatcmpl-$REQUEST_ID","object":"chat.completion.chunk","created":1,"model":"$MODEL","choices":[{"index":0,"delta":{"role":"assistant","content":"ok"},"finish_reason":null}]}\n\n',
'data: {"id":"chatcmpl-$REQUEST_ID","object":"chat.completion.chunk","created":1,"model":"$MODEL","choices":[],"usage":{"prompt_tokens":1840,"completion_tokens":412,"total_tokens":2252}}\n\n',
"data: [DONE]\n\n",
),
),
expected=ExactExpected(
spend=0.003234, input_cost=0.00138, output_cost=0.001854, prompt_tokens=1840, completion_tokens=412
),
)
# azure_ai/grok-4.6 (Grok on Azure AI Foundry reports reasoning_tokens outside completion_tokens, like xai)
AZURE_AI_GROK_4_6_REASONING_FOLDED_INTO_COMPLETION: Final = CostTrackingTestCase(
name="azure-ai-grok-4.6-reasoning_folded_into_completion",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/grok-4.6",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "azure-ai-grok-4.6-reasoning"}],
"stream": False,
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "grok-4.6",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "ok"}, "finish_reason": "stop"}],
"usage": {
"prompt_tokens": 1840,
"completion_tokens": 412,
"total_tokens": 2552,
"completion_tokens_details": {"reasoning_tokens": 300},
},
},
),
expected=ExactExpected(
spend=0.0108, input_cost=0.00368, output_cost=0.00712, prompt_tokens=1840, completion_tokens=712
),
)
AZURE_AI_GROK_4_6_REASONING_FOLDED_INTO_COMPLETION_STREAM: Final = CostTrackingTestCase(
name="azure-ai-grok-4.6-reasoning_folded_into_completion-stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="azure_ai/grok-4.6",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "azure-ai-grok-4.6-reasoning-stream"}],
"stream": True,
"stream_options": {"include_usage": True},
},
response=SseResponse(
content_type="text/event-stream",
frames=(
'data: {"id":"chatcmpl-$REQUEST_ID","object":"chat.completion.chunk","created":1,"model":"grok-4.6","choices":[{"index":0,"delta":{"role":"assistant","content":"ok"},"finish_reason":null}]}\n\n',
'data: {"id":"chatcmpl-$REQUEST_ID","object":"chat.completion.chunk","created":1,"model":"grok-4.6","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}\n\n',
'data: {"id":"chatcmpl-$REQUEST_ID","object":"chat.completion.chunk","created":1,"model":"grok-4.6","choices":[],"usage":{"prompt_tokens":1840,"completion_tokens":412,"total_tokens":2552,"completion_tokens_details":{"reasoning_tokens":300}}}\n\n',
"data: [DONE]\n\n",
),
),
expected=ExactExpected(
spend=0.0108, input_cost=0.00368, output_cost=0.00712, prompt_tokens=1840, completion_tokens=712
),
)
# azure_ai/claude-haiku-4-5 (Azure AI Foundry Anthropic, web search billed per server_tool_use.web_search_requests)
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_WEB_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-haiku-4-5-messages_web_search_two_requests",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-haiku-4-5",
endpoint="/v1/messages",
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [{"role": "user", "content": "azure-ai-claude-haiku-4-5-messages_web_search_two_requests"}],
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$UNIQUE_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-4-5",
"content": [{"type": "text", "text": "scripted answer a1b2c3d4e5f6"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
rollups=True,
),
)
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_WEB_SEARCH_STREAM: Final = CostTrackingTestCase(
name="azure-ai-claude-haiku-4-5-messages_web_search_two_requests_stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="azure_ai/claude-haiku-4-5",
endpoint="/v1/messages",
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [{"role": "user", "content": "azure-ai-claude-haiku-4-5-messages_web_search_two_requests_stream"}],
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
"stream": True,
},
response=SseResponse(
content_type="text/event-stream",
frames=(
'event: message_start\ndata: {"type":"message_start","message":{"id":"msg_$REQUEST_ID","type":"message","role":"assistant","model":"claude-haiku-4-5","content":[],"stop_reason":null,"stop_sequence":null,"usage":{"input_tokens":1840}}}',
'event: content_block_start\ndata: {"type":"content_block_start","index":0,"content_block":{"type":"text","text":""}}',
'event: content_block_delta\ndata: {"type":"content_block_delta","index":0,"delta":{"type":"text_delta","text":"scripted response"}}',
'event: content_block_stop\ndata: {"type":"content_block_stop","index":0}',
'event: message_delta\ndata: {"type":"message_delta","delta":{"stop_reason":"end_turn"},"usage":{"output_tokens":412,"server_tool_use":{"web_search_requests":2}}}',
'event: message_stop\ndata: {"type":"message_stop"}',
),
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
),
)
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_WEB_SEARCH_ONE_REQUEST: Final = CostTrackingTestCase(
name="azure-ai-claude-haiku-4-5-messages_web_search_one_request",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-haiku-4-5",
endpoint="/v1/messages",
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [{"role": "user", "content": "azure-ai-claude-haiku-4-5-messages_web_search_one_request"}],
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-4-5",
"content": [{"type": "text", "text": "scripted answer 0f1e2d3c4b5a"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 1}},
},
),
expected=ExactExpected(
spend=0.0139,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.01,
),
)
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_NO_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-haiku-4-5-messages_no_server_tool_use",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-haiku-4-5",
endpoint="/v1/messages",
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [{"role": "user", "content": "azure-ai-claude-haiku-4-5-messages_no_server_tool_use"}],
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-4-5",
"content": [{"type": "text", "text": "scripted answer 9a8b7c6d5e4f"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412},
},
),
expected=ExactExpected(
spend=0.0039,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.0,
),
)
AZURE_AI_CLAUDE_HAIKU_4_5_CHAT_WEB_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-haiku-4-5-chat_web_search_three_requests",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-haiku-4-5",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "azure-ai-claude-haiku-4-5-chat_web_search_three_requests"}],
"stream": False,
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 5}],
"allowed_openai_params": ["web_search_options"],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-4-5",
"content": [{"type": "text", "text": "scripted answer 5e4f3a2b1c0d"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 3}},
},
),
expected=ExactExpected(
spend=0.0339,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.03,
),
)
AZURE_AI_CLAUDE_HAIKU_4_5_CHAT_WEB_SEARCH_STREAM: Final = CostTrackingTestCase(
name="azure-ai-claude-haiku-4-5-chat_web_search_two_requests_stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="azure_ai/claude-haiku-4-5",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "azure-ai-claude-haiku-4-5-chat_web_search_two_requests_stream"}],
"stream": True,
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 5}],
"allowed_openai_params": ["web_search_options"],
},
response=SseResponse(
content_type="text/event-stream",
frames=sse_frames(
{
"type": "message_start",
"message": {
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-4-5",
"content": [],
"stop_reason": None,
"usage": {"input_tokens": 1840},
},
},
{"type": "content_block_start", "index": 0, "content_block": {"type": "text", "text": ""}},
{
"type": "content_block_delta",
"index": 0,
"delta": {"type": "text_delta", "text": "scripted answer 2c3d4e5f6a7b"},
},
{"type": "content_block_stop", "index": 0},
{
"type": "message_delta",
"delta": {"stop_reason": "end_turn"},
"usage": {"output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
{"type": "message_stop"},
),
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
),
)
AZURE_AI_CLAUDE_HAIKU_4_5_RESPONSES_WEB_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-haiku-4-5-responses_web_search_two_requests",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-haiku-4-5",
endpoint="/v1/responses",
request={
"model": "$MODEL",
"input": "azure-ai-claude-haiku-4-5-responses_web_search_two_requests",
"tools": [{"type": "web_search"}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-4-5",
"content": [{"type": "text", "text": "scripted answer 7d6e5f4a3b2c"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
),
)
# azure_ai/claude-sonnet-4-5 (public row has no search price, the deployment owns input, output and search rates)
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_WEB_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-sonnet-4-5-deployment_rates_messages_web_search_two_requests",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-sonnet-4-5",
endpoint="/v1/messages",
deployment=Deployment(input_cost_per_token=1e-06, output_cost_per_token=5e-06, search_cost_per_query=0.01),
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [
{"role": "user", "content": "azure-ai-claude-sonnet-4-5-deployment_rates_messages_web_search_two_requests"}
],
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$UNIQUE_ID",
"type": "message",
"role": "assistant",
"model": "claude-sonnet-4-5",
"content": [{"type": "text", "text": "scripted answer b1c2d3e4f5a6"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
rollups=True,
),
)
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_WEB_SEARCH_STREAM: Final = CostTrackingTestCase(
name="azure-ai-claude-sonnet-4-5-deployment_rates_messages_web_search_two_requests_stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="azure_ai/claude-sonnet-4-5",
endpoint="/v1/messages",
deployment=Deployment(input_cost_per_token=1e-06, output_cost_per_token=5e-06, search_cost_per_query=0.01),
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [
{
"role": "user",
"content": "azure-ai-claude-sonnet-4-5-deployment_rates_messages_web_search_two_requests_stream",
}
],
"stream": True,
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
},
response=SseResponse(
content_type="text/event-stream",
frames=sse_frames(
{
"type": "message_start",
"message": {
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-sonnet-4-5",
"content": [],
"stop_reason": None,
"usage": {"input_tokens": 1840},
},
},
{"type": "content_block_start", "index": 0, "content_block": {"type": "text", "text": ""}},
{
"type": "content_block_delta",
"index": 0,
"delta": {"type": "text_delta", "text": "scripted answer c2d3e4f5a6b1"},
},
{"type": "content_block_stop", "index": 0},
{
"type": "message_delta",
"delta": {"stop_reason": "end_turn"},
"usage": {"output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
{"type": "message_stop"},
),
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
),
)
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_WEB_SEARCH_ZERO_RATE: Final = CostTrackingTestCase(
name="azure-ai-claude-sonnet-4-5-deployment_zero_search_rate_messages_two_requests",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-sonnet-4-5",
endpoint="/v1/messages",
deployment=Deployment(input_cost_per_token=1e-06, output_cost_per_token=5e-06, search_cost_per_query=0.0),
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [
{"role": "user", "content": "azure-ai-claude-sonnet-4-5-deployment_zero_search_rate_messages_two_requests"}
],
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-sonnet-4-5",
"content": [{"type": "text", "text": "scripted answer d3e4f5a6b1c2"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
),
expected=ExactExpected(
spend=0.0039,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.0,
),
)
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_NO_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-sonnet-4-5-deployment_rates_messages_no_server_tool_use",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-sonnet-4-5",
endpoint="/v1/messages",
deployment=Deployment(input_cost_per_token=1e-06, output_cost_per_token=5e-06, search_cost_per_query=0.01),
request={
"model": "$MODEL",
"max_tokens": 412,
"messages": [
{"role": "user", "content": "azure-ai-claude-sonnet-4-5-deployment_rates_messages_no_server_tool_use"}
],
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-sonnet-4-5",
"content": [{"type": "text", "text": "scripted answer e4f5a6b1c2d3"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412},
},
),
expected=ExactExpected(
spend=0.0039,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.0,
),
)
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_CHAT_WEB_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-sonnet-4-5-deployment_rates_chat_web_search_three_requests",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-sonnet-4-5",
deployment=Deployment(input_cost_per_token=1e-06, output_cost_per_token=5e-06, search_cost_per_query=0.01),
request={
"model": "$MODEL",
"messages": [
{"role": "user", "content": "azure-ai-claude-sonnet-4-5-deployment_rates_chat_web_search_three_requests"}
],
"stream": False,
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 5}],
"allowed_openai_params": ["web_search_options"],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-sonnet-4-5",
"content": [{"type": "text", "text": "scripted answer f5a6b1c2d3e4"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 3}},
},
),
expected=ExactExpected(
spend=0.0339,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.03,
),
)
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_CHAT_WEB_SEARCH_STREAM: Final = CostTrackingTestCase(
name="azure-ai-claude-sonnet-4-5-deployment_rates_chat_web_search_two_requests_stream",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="azure_ai/claude-sonnet-4-5",
deployment=Deployment(input_cost_per_token=1e-06, output_cost_per_token=5e-06, search_cost_per_query=0.01),
request={
"model": "$MODEL",
"messages": [
{
"role": "user",
"content": "azure-ai-claude-sonnet-4-5-deployment_rates_chat_web_search_two_requests_stream",
}
],
"stream": True,
"tools": [{"type": "web_search_20250305", "name": "web_search", "max_uses": 3}],
"allowed_openai_params": ["web_search_options"],
},
response=SseResponse(
content_type="text/event-stream",
frames=sse_frames(
{
"type": "message_start",
"message": {
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-sonnet-4-5",
"content": [],
"stop_reason": None,
"usage": {"input_tokens": 1840},
},
},
{"type": "content_block_start", "index": 0, "content_block": {"type": "text", "text": ""}},
{
"type": "content_block_delta",
"index": 0,
"delta": {"type": "text_delta", "text": "scripted answer a6b1c2d3e4f5"},
},
{"type": "content_block_stop", "index": 0},
{
"type": "message_delta",
"delta": {"stop_reason": "end_turn"},
"usage": {"output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
{"type": "message_stop"},
),
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
),
)
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_RESPONSES_WEB_SEARCH: Final = CostTrackingTestCase(
name="azure-ai-claude-sonnet-4-5-deployment_rates_responses_web_search_two_requests",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="azure_ai/claude-sonnet-4-5",
endpoint="/v1/responses",
deployment=Deployment(input_cost_per_token=1e-06, output_cost_per_token=5e-06, search_cost_per_query=0.01),
request={
"model": "$MODEL",
"input": "azure-ai-claude-sonnet-4-5-deployment_rates_responses_web_search_two_requests",
"tools": [{"type": "web_search"}],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-sonnet-4-5",
"content": [{"type": "text", "text": "scripted answer b1c2d3e4f5a7"}],
"stop_reason": "end_turn",
"usage": {"input_tokens": 1840, "output_tokens": 412, "server_tool_use": {"web_search_requests": 2}},
},
),
expected=ExactExpected(
spend=0.0239,
input_cost=0.00184,
output_cost=0.00206,
prompt_tokens=1840,
completion_tokens=412,
tool_usage_cost=0.02,
),
)
# groq/qwen/qwen3.8-27b
GROQ_QWEN_3_8_JSON: Final = CostTrackingTestCase(
name="groq-qwen-3.8-json",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="groq/qwen/qwen3.8-27b",
request={"model": "$MODEL", "messages": [{"role": "user", "content": "groq-qwen-3.8-json"}], "stream": False},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "$MODEL",
"choices": [
{"index": 0, "message": {"role": "assistant", "content": "scripted response"}, "finish_reason": "stop"}
],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
"service_tier": "default",
"x_groq": {"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252}},
},
),
expected=ExactExpected(
spend=0.00312, input_cost=0.001472, output_cost=0.001648, prompt_tokens=1840, completion_tokens=412
),
)
GROQ_QWEN_3_8_STREAM_X_GROQ_RECOUNT: Final = CostTrackingTestCase(
name="groq-qwen-3.8-stream_x_groq_recount",
covers="quota_management.spend_tracking.scripted_wire.logs_cost",
model="groq/qwen/qwen3.8-27b",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "groq-qwen-3.8-stream_x_groq_recount"}],
"stream": True,
},
response=SseResponse(
content_type="text/event-stream",
frames=(
'data: {"id":"chatcmpl-$REQUEST_ID","object":"chat.completion.chunk","created":1,"model":"$MODEL","choices":[{"index":0,"delta":{"role":"assistant","content":"ok"},"finish_reason":null}]}\n\n',
'data: {"id":"chatcmpl-$REQUEST_ID","object":"chat.completion.chunk","created":1,"model":"$MODEL","choices":[],"x_groq":{"usage":{"prompt_tokens":1840,"completion_tokens":412,"total_tokens":2252}}}\n\n',
"data: [DONE]\n\n",
),
),
expected=RecountExpected(recount=RecountRates(input_cost_per_token=8e-07, output_cost_per_token=4e-06)),
)
# cohere_chat/v2/command-a-03-2025
COHERE_COMMAND_A_V2_TOKENS: Final = CostTrackingTestCase(
name="cohere-command-a-v2-tokens",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="cohere_chat/v2/command-a-03-2025",
deployment=Deployment(model="cohere_chat/v2/command-a-03-2025"),
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "cohere-command-a-v2-tokens"}],
"stream": False,
},
response=JsonResponse(
content_type="application/json",
body={
"id": "$REQUEST_ID",
"message": {"role": "assistant", "content": [{"type": "text", "text": "scripted response"}]},
"finish_reason": "COMPLETE",
"usage": {
"tokens": {"input_tokens": 1840, "output_tokens": 412, "total_tokens": 2252},
"billed_units": {"input_tokens": 1800, "output_tokens": 400},
},
},
),
expected=ExactExpected(
spend=0.00874252, input_cost=0.0046184, output_cost=0.00412412, prompt_tokens=1840, completion_tokens=412
),
)
# mistral/mistral-medium-2604
MISTRAL_MEDIUM_2604_JSON: Final = CostTrackingTestCase(
name="mistral-medium-2604-json",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="mistral/mistral-medium-2604",
request={"model": "$MODEL", "messages": [{"role": "user", "content": "mistral-medium-2604-json"}], "stream": False},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "$MODEL",
"choices": [
{"index": 0, "message": {"role": "assistant", "content": "scripted response"}, "finish_reason": "stop"}
],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
"service_tier": "default",
},
),
expected=ExactExpected(
spend=0.00587252, input_cost=0.0027784, output_cost=0.00309412, prompt_tokens=1840, completion_tokens=412
),
)
# xai/grok-4.7
XAI_GROK_4_7_INPUT_TEXT: Final = CostTrackingTestCase(
name="xai-grok-4.7-input_text",
covers="quota_management.spend_tracking.cost_matrix.logs_cost",
model="xai/grok-4.7",
request={
"model": "$MODEL",
"messages": [{"role": "user", "content": "Say hello."}],
"stream": False,
"allowed_openai_params": [],
},
response=JsonResponse(
content_type="application/json",
body={
"id": "chatcmpl-$REQUEST_ID",
"object": "chat.completion",
"model": "grok-4.7",
"choices": [{"index": 0, "message": {"role": "assistant", "content": "Hello."}, "finish_reason": "stop"}],
"usage": {"prompt_tokens": 1840, "completion_tokens": 412, "total_tokens": 2252},
},
),
expected=ExactExpected(
spend=0.006152, input_cost=0.00368, output_cost=0.002472, prompt_tokens=1840, completion_tokens=412
),
)
AZURE_AI_CLAUDE_FABLE_5_1_INPUT_TEXT: Final = CLAUDE_FABLE_5_1_INPUT_TEXT.model_copy(
update={
"name": "azure_ai/claude-fable-5-1-input_text",
"model": "azure_ai/claude-fable-5-1",
"response": JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-fable-5-1",
"content": [
{"type": "thinking", "thinking": "scripted thought", "signature": "scripted signature"},
{"type": "text", "text": "scripted answer 3d584fa07ae1"},
],
"stop_reason": "end_turn",
"stop_sequence": None,
"stop_details": None,
"container": None,
"diagnostics": None,
"usage": {
"input_tokens": 1840,
"cache_creation_input_tokens": 0,
"cache_read_input_tokens": 0,
"cache_creation": {"ephemeral_5m_input_tokens": 0, "ephemeral_1h_input_tokens": 0},
"output_tokens": 412,
"output_tokens_details": {"thinking_tokens": 0},
"service_tier": "standard",
"inference_geo": "not_available",
"speed": "standard",
},
},
),
"expected": ExactExpected(
spend=0.0663, input_cost=0.03128, output_cost=0.03502, prompt_tokens=1840, completion_tokens=412
),
}
)
AZURE_AI_CLAUDE_FABLE_5_1_STREAM: Final = CLAUDE_FABLE_5_1_STREAM.model_copy(
update={
"name": "azure_ai/claude-fable-5-1-stream",
"model": "azure_ai/claude-fable-5-1",
"response": SseResponse(
content_type="text/event-stream",
frames=sse_frames(
*(
{
"type": "message_start",
"message": {
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-fable-5-1",
"content": [],
"stop_reason": None,
"stop_sequence": None,
"stop_details": None,
"container": None,
"diagnostics": None,
"usage": {
"input_tokens": 1840,
"output_tokens": 1,
"cache_creation_input_tokens": 0,
"cache_read_input_tokens": 0,
"cache_creation": {"ephemeral_5m_input_tokens": 0, "ephemeral_1h_input_tokens": 0},
"service_tier": "standard",
"inference_geo": "not_available",
"speed": "standard",
},
},
},
{"type": "content_block_start", "index": 0, "content_block": {"type": "thinking", "thinking": "", "signature": ""}},
{"type": "ping"},
{"type": "content_block_delta", "index": 0, "delta": {"type": "thinking_delta", "thinking": ""}},
{
"type": "content_block_delta",
"index": 0,
"delta": {"type": "signature_delta", "signature": "scripted signature"},
},
{"type": "content_block_stop", "index": 0},
{"type": "content_block_start", "index": 1, "content_block": {"type": "text", "text": ""}},
*(
{
"type": "content_block_delta",
"index": 1,
"delta": {"type": "text_delta", "text": text},
}
for text in (
"scripted ",
"answer ",
"3d",
"58",
"4f",
"a0",
"7a",
"e1",
"",
"",
)
),
{"type": "content_block_stop", "index": 1},
{
"type": "message_delta",
"delta": {
"container": None,
"stop_details": None,
"stop_reason": "end_turn",
"stop_sequence": None,
},
"usage": {
"input_tokens": 1840,
"output_tokens": 412,
"cache_creation_input_tokens": 0,
"cache_read_input_tokens": 0,
"output_tokens_details": {"thinking_tokens": 0},
},
},
{"type": "message_stop"},
)
),
),
"expected": ExactExpected(
spend=0.0663, input_cost=0.03128, output_cost=0.03502, prompt_tokens=1840, completion_tokens=412
),
}
)
AZURE_AI_CLAUDE_HAIKU_5_5_INPUT_TEXT: Final = CLAUDE_HAIKU_5_5_INPUT_TEXT.model_copy(
update={
"name": "azure_ai/claude-haiku-5-5-input_text",
"model": "azure_ai/claude-haiku-5-5",
"response": JsonResponse(
content_type="application/json",
body={
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-5-5",
"content": [
{"type": "thinking", "thinking": "scripted thought", "signature": "scripted signature"},
{"type": "text", "text": "scripted answer 10673f3a1db3"},
],
"container": None,
"stop_reason": "end_turn",
"stop_sequence": None,
"stop_details": None,
"usage": {
"input_tokens": 1840,
"cache_creation_input_tokens": 0,
"cache_read_input_tokens": 0,
"cache_creation": {"ephemeral_5m_input_tokens": 0, "ephemeral_1h_input_tokens": 0},
"output_tokens": 412,
"output_tokens_details": {"thinking_tokens": 0},
"service_tier": "standard",
"inference_geo": "global",
"speed": "standard",
},
"diagnostics": None,
},
),
"expected": ExactExpected(
spend=0.000741, input_cost=0.0003496, output_cost=0.0003914, prompt_tokens=1840, completion_tokens=412
),
}
)
AZURE_AI_CLAUDE_HAIKU_5_5_STREAM: Final = CLAUDE_HAIKU_5_5_STREAM.model_copy(
update={
"name": "azure_ai/claude-haiku-5-5-stream",
"model": "azure_ai/claude-haiku-5-5",
"response": SseResponse(
content_type="text/event-stream",
frames=sse_frames(
*(
{
"type": "message_start",
"message": {
"id": "msg_$REQUEST_ID",
"type": "message",
"role": "assistant",
"model": "claude-haiku-5-5",
"content": [],
"stop_reason": None,
"stop_sequence": None,
"stop_details": None,
"container": None,
"diagnostics": None,
"usage": {
"input_tokens": 1840,
"output_tokens": 1,
"cache_creation_input_tokens": 0,
"cache_read_input_tokens": 0,
"cache_creation": {
"ephemeral_5m_input_tokens": 0,
"ephemeral_1h_input_tokens": 0,
},
"service_tier": "standard",
"inference_geo": "global",
"speed": "standard",
},
},
},
{
"type": "content_block_start",
"index": 0,
"content_block": {"type": "thinking", "thinking": "", "signature": ""},
},
{"type": "ping"},
{
"type": "content_block_delta",
"index": 0,
"delta": {"type": "thinking_delta", "thinking": ""},
},
{
"type": "content_block_delta",
"index": 0,
"delta": {"type": "signature_delta", "signature": "scripted signature"},
},
{"type": "content_block_stop", "index": 0},
{
"type": "content_block_start",
"index": 1,
"content_block": {"type": "text", "text": ""},
},
*(
{
"type": "content_block_delta",
"index": 1,
"delta": {"type": "text_delta", "text": text},
}
for text in (
"scripted ",
"ans",
"wer ",
"2",
"b",
"f",
"9",
"b",
"c",
"b",
"a",
"4",
"444",
)
),
{"type": "content_block_stop", "index": 1},
{
"type": "message_delta",
"delta": {
"container": None,
"stop_details": None,
"stop_reason": "end_turn",
"stop_sequence": None,
},
"usage": {
"input_tokens": 1840,
"cache_creation_input_tokens": 0,
"cache_read_input_tokens": 0,
"output_tokens": 412,
"output_tokens_details": {"thinking_tokens": 0},
},
},
{"type": "message_stop"},
)
),
),
"expected": ExactExpected(
spend=0.000741, input_cost=0.0003496, output_cost=0.0003914, prompt_tokens=1840, completion_tokens=412
),
}
)
PERPLEXITY_SONAR_NEXT_OFF_PEAK: Final = PERPLEXITY_SONAR_NEXT_NO_SEARCH.model_copy(
update={
"name": "perplexity-sonar-next-off_peak",
"deployment": Deployment(
off_peak_pricing=OffPeakPricing(
hours_utc=ALWAYS_OPEN_OFF_PEAK_WINDOW, input_cost_per_token=2.26e-06, output_cost_per_token=2.1e-06
)
),
"expected": ExactExpected(
spend=0.0050236, input_cost=0.0041584, output_cost=0.0008652, prompt_tokens=1840, completion_tokens=412
),
}
)
AZURE_AI_CLAUDE_FABLE_5_1_CACHE_READ: Final = _azure_ai_claude_shape_case(
CLAUDE_FABLE_5_1_CACHE_READ,
"azure_ai/claude-fable-5-1-cache_read",
"azure_ai/claude-fable-5-1",
_azure_ai_claude_usage(input_tokens=14, output_tokens=6, inference_geo="not_available", cache_read_input_tokens=27003),
ExactExpected(
spend=0.012224275,
input_cost=0.011714275,
output_cost=0.00051,
prompt_tokens=27017,
completion_tokens=6,
),
)
AZURE_AI_CLAUDE_FABLE_5_1_CACHE_WRITE_5M: Final = _azure_ai_claude_shape_case(
CLAUDE_FABLE_5_1_CACHE_WRITE_5M,
"azure_ai/claude-fable-5-1-cache_write_5m",
"azure_ai/claude-fable-5-1",
_azure_ai_claude_usage(
input_tokens=14,
output_tokens=5,
inference_geo="not_available",
cache_creation_input_tokens=27011,
ephemeral_5m_input_tokens=27011,
),
ExactExpected(
spend=0.57464675,
input_cost=0.57422175,
output_cost=0.000425,
prompt_tokens=27025,
completion_tokens=5,
),
)
AZURE_AI_CLAUDE_FABLE_5_1_CACHE_WRITE_1H: Final = _azure_ai_claude_shape_case(
CLAUDE_FABLE_5_1_CACHE_WRITE_1H,
"azure_ai/claude-fable-5-1-cache_write_1h",
"azure_ai/claude-fable-5-1",
_azure_ai_claude_usage(
input_tokens=14,
output_tokens=5,
inference_geo="not_available",
cache_creation_input_tokens=27011,
ephemeral_1h_input_tokens=27011,
),
ExactExpected(
spend=0.919037,
input_cost=0.918612,
output_cost=0.000425,
prompt_tokens=27025,
completion_tokens=5,
),
)
AZURE_AI_CLAUDE_FABLE_5_1_THINKING: Final = _azure_ai_claude_shape_case(
CLAUDE_FABLE_5_1_THINKING,
"azure_ai/claude-fable-5-1-thinking",
"azure_ai/claude-fable-5-1",
_azure_ai_claude_usage(input_tokens=18, output_tokens=99, inference_geo="not_available", thinking_tokens=33),
ExactExpected(
spend=0.008721, input_cost=0.000306, output_cost=0.008415, prompt_tokens=18, completion_tokens=99
),
[
{"type": "thinking", "thinking": "scripted thought", "signature": "scripted signature"},
{"type": "text", "text": "scripted answer for cost fixtures"},
],
adaptive_thinking=True,
)
AZURE_AI_CLAUDE_FABLE_5_1_TOOL_CALL: Final = _azure_ai_claude_shape_case(
CLAUDE_FABLE_5_1_TOOL_CALL,
"azure_ai/claude-fable-5-1-tool_call",
"azure_ai/claude-fable-5-1",
_azure_ai_claude_usage(input_tokens=385, output_tokens=50, inference_geo="not_available"),
ExactExpected(
spend=0.010795, input_cost=0.006545, output_cost=0.00425, prompt_tokens=385, completion_tokens=50
),
)
AZURE_AI_CLAUDE_HAIKU_5_5_CACHE_READ: Final = _azure_ai_claude_shape_case(
CLAUDE_HAIKU_5_5_CACHE_READ,
"azure_ai/claude-haiku-5-5-cache_read",
"azure_ai/claude-haiku-5-5",
_azure_ai_claude_usage(input_tokens=14, output_tokens=5, inference_geo="global", cache_read_input_tokens=27003),
ExactExpected(
spend=0.000520467,
input_cost=0.000515717,
output_cost=0.00000475,
prompt_tokens=27017,
completion_tokens=5,
),
)
AZURE_AI_CLAUDE_HAIKU_5_5_CACHE_WRITE_5M: Final = _azure_ai_claude_shape_case(
CLAUDE_HAIKU_5_5_CACHE_WRITE_5M,
"azure_ai/claude-haiku-5-5-cache_write_5m",
"azure_ai/claude-haiku-5-5",
_azure_ai_claude_usage(
input_tokens=14,
output_tokens=6,
inference_geo="global",
cache_creation_input_tokens=27011,
ephemeral_5m_input_tokens=27011,
),
ExactExpected(
spend=0.0064234725,
input_cost=0.0064177725,
output_cost=0.0000057,
prompt_tokens=27025,
completion_tokens=6,
),
)
AZURE_AI_CLAUDE_HAIKU_5_5_CACHE_WRITE_1H: Final = _azure_ai_claude_shape_case(
CLAUDE_HAIKU_5_5_CACHE_WRITE_1H,
"azure_ai/claude-haiku-5-5-cache_write_1h",
"azure_ai/claude-haiku-5-5",
_azure_ai_claude_usage(
input_tokens=14,
output_tokens=6,
inference_geo="global",
cache_creation_input_tokens=27011,
ephemeral_1h_input_tokens=27011,
),
ExactExpected(
spend=0.01027254,
input_cost=0.01026684,
output_cost=0.0000057,
prompt_tokens=27025,
completion_tokens=6,
),
)
AZURE_AI_CLAUDE_HAIKU_5_5_THINKING: Final = _azure_ai_claude_shape_case(
CLAUDE_HAIKU_5_5_THINKING,
"azure_ai/claude-haiku-5-5-thinking",
"azure_ai/claude-haiku-5-5",
_azure_ai_claude_usage(input_tokens=18, output_tokens=95, inference_geo="global", thinking_tokens=54),
ExactExpected(
spend=0.00009367, input_cost=0.00000342, output_cost=0.00009025, prompt_tokens=18, completion_tokens=95
),
[
{"type": "thinking", "thinking": "scripted thought", "signature": "scripted signature"},
{"type": "text", "text": "scripted answer for cost fixtures"},
],
adaptive_thinking=True,
)
AZURE_AI_CLAUDE_HAIKU_5_5_TOOL_CALL: Final = _azure_ai_claude_shape_case(
CLAUDE_HAIKU_5_5_TOOL_CALL,
"azure_ai/claude-haiku-5-5-tool_call",
"azure_ai/claude-haiku-5-5",
_azure_ai_claude_usage(input_tokens=511, output_tokens=34, inference_geo="global"),
ExactExpected(
spend=0.00012939, input_cost=0.00009709, output_cost=0.0000323, prompt_tokens=511, completion_tokens=34
),
)
AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_INPUT_TEXT: Final = _azure_ai_claude_messages_case(
AZURE_AI_CLAUDE_FABLE_5_1_INPUT_TEXT,
CLAUDE_FABLE_5_1_MESSAGES_INPUT_TEXT,
"azure_ai/claude-fable-5-1-messages_input_text",
"azure_ai/claude-fable-5-1",
)
AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_STREAM: Final = _azure_ai_claude_messages_case(
AZURE_AI_CLAUDE_FABLE_5_1_STREAM,
CLAUDE_FABLE_5_1_MESSAGES_STREAM,
"azure_ai/claude-fable-5-1-messages_stream",
"azure_ai/claude-fable-5-1",
)
AZURE_AI_CLAUDE_FABLE_5_1_ROUTE_BRIDGES: Final = bridge_pairs(
(AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_INPUT_TEXT, AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_STREAM),
("/v1/responses",),
)
AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_INPUT_TEXT: Final = _azure_ai_claude_messages_case(
AZURE_AI_CLAUDE_HAIKU_5_5_INPUT_TEXT,
CLAUDE_HAIKU_5_5_MESSAGES_INPUT_TEXT,
"azure_ai/claude-haiku-5-5-messages_input_text",
"azure_ai/claude-haiku-5-5",
)
AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_STREAM: Final = _azure_ai_claude_messages_case(
AZURE_AI_CLAUDE_HAIKU_5_5_STREAM,
CLAUDE_HAIKU_5_5_MESSAGES_STREAM,
"azure_ai/claude-haiku-5-5-messages_stream",
"azure_ai/claude-haiku-5-5",
)
AZURE_AI_CLAUDE_HAIKU_5_5_ROUTE_BRIDGES: Final = bridge_pairs(
(AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_INPUT_TEXT, AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_STREAM),
("/v1/responses",),
)
AZURE_AI_CLAUDE_HAIKU_5_5_TIERED_INPUT_ABOVE_100K: Final = with_usage(
AZURE_AI_CLAUDE_HAIKU_5_5_INPUT_TEXT,
"azure_ai/claude-haiku-5-5-tiered_input_above_100k",
{
"input_tokens": 150000,
"output_tokens": 480,
"cache_read_input_tokens": 0,
"cache_creation_input_tokens": 0,
"cache_creation": {"ephemeral_5m_input_tokens": 0, "ephemeral_1h_input_tokens": 0},
},
ExactExpected(
spend=0.14478, input_cost=0.1425, output_cost=0.00228, prompt_tokens=150000, completion_tokens=480
),
)
AZURE_AI_CLAUDE_HAIKU_5_5_TIERED_CACHE_READ_ABOVE_100K: Final = with_usage(
AZURE_AI_CLAUDE_HAIKU_5_5_CACHE_READ,
"azure_ai/claude-haiku-5-5-tiered_cache_read_above_100k",
{
"input_tokens": 4096,
"output_tokens": 480,
"cache_read_input_tokens": 110592,
"cache_creation_input_tokens": 0,
"cache_creation": {"ephemeral_5m_input_tokens": 0, "ephemeral_1h_input_tokens": 0},
},
ExactExpected(
spend=0.01667744, input_cost=0.01439744, output_cost=0.00228, prompt_tokens=114688, completion_tokens=480
),
)
CASES: Final[tuple[CostTrackingTestCase, ...]] = (
PERPLEXITY_SONAR_NEXT_OFF_PEAK,
NOVA_NEXT_TRANSCRIPTIONS_PER_SECOND,
AMAZON_NOVA_CANVAS_NEXT_IMAGES_ONE,
COHERE_EMBEDDINGS_V5,
BEDROCK_EMBEDDINGS_TITAN_V2,
BEDROCK_COHERE_EMBEDDINGS_V4,
COHERE_RERANK_V4_ONE,
COHERE_RERANK_V4_THREE,
COHERE_RERANK_V4_TOTAL_TOKENS_FALLBACK,
BEDROCK_COHERE_RERANK_V4,
TEXT_COMPLETIONS_OPENAI_BASIC,
TEXT_COMPLETIONS_OPENAI_STREAM_USAGE,
TEXT_COMPLETIONS_OPENAI_N_BEST,
DASHSCOPE_QWEN4_MAX_TIERED_INPUT,
DASHSCOPE_QWEN4_MAX_TIERED_BOUNDARY_STAYS_LOWER_TIER,
DASHSCOPE_QWEN4_MAX_TIERED_SECOND_TIER,
DASHSCOPE_QWEN4_MAX_TIERED_ABOVE_TOP_RANGE,
OPENROUTER_ANTHROPIC_CLAUDE_SONNET_5_PROVIDER_REPORTED_COST,
OPENROUTER_ANTHROPIC_CLAUDE_SONNET_5_TOKEN_PRICED,
PERPLEXITY_SONAR_NEXT_NO_SEARCH,
DEEPSEEK_DEEPSEEK_V4_CHAT_PROMPT_CACHE_HIT,
DEEPSEEK_DEEPSEEK_V4_CHAT_NO_CACHE_FIELDS_BILLS_ZERO_CACHE,
XAI_GROK_5_REASONING_FOLDED_INTO_COMPLETION,
XAI_GROK_5_LIVE_SEARCH,
XAI_GROK_5_PROVIDER_REPORTED_COST,
BEDROCK_INVOKE_HAIKU_JSON,
BEDROCK_INVOKE_HAIKU_STREAM,
BEDROCK_INVOKE_HAIKU_CHAT_STREAM,
BEDROCK_INVOKE_HAIKU_RESPONSES,
AZURE_AI_GPT_5_4_MINI_LATEST,
AZURE_AI_GPT_5_4_MINI_LATEST_STREAM,
AZURE_AI_GROK_4_6_REASONING_FOLDED_INTO_COMPLETION,
AZURE_AI_GROK_4_6_REASONING_FOLDED_INTO_COMPLETION_STREAM,
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_WEB_SEARCH,
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_WEB_SEARCH_STREAM,
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_WEB_SEARCH_ONE_REQUEST,
AZURE_AI_CLAUDE_HAIKU_4_5_MESSAGES_NO_SEARCH,
AZURE_AI_CLAUDE_HAIKU_4_5_CHAT_WEB_SEARCH,
AZURE_AI_CLAUDE_HAIKU_4_5_CHAT_WEB_SEARCH_STREAM,
AZURE_AI_CLAUDE_HAIKU_4_5_RESPONSES_WEB_SEARCH,
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_WEB_SEARCH,
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_WEB_SEARCH_STREAM,
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_WEB_SEARCH_ZERO_RATE,
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_MESSAGES_NO_SEARCH,
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_CHAT_WEB_SEARCH,
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_CHAT_WEB_SEARCH_STREAM,
AZURE_AI_CLAUDE_SONNET_4_5_DEPLOYMENT_RESPONSES_WEB_SEARCH,
GROQ_QWEN_3_8_JSON,
GROQ_QWEN_3_8_STREAM_X_GROQ_RECOUNT,
COHERE_COMMAND_A_V2_TOKENS,
MISTRAL_MEDIUM_2604_JSON,
XAI_GROK_4_7_INPUT_TEXT,
AZURE_AI_CLAUDE_FABLE_5_1_INPUT_TEXT,
AZURE_AI_CLAUDE_FABLE_5_1_STREAM,
AZURE_AI_CLAUDE_FABLE_5_1_CACHE_READ,
AZURE_AI_CLAUDE_FABLE_5_1_CACHE_WRITE_5M,
AZURE_AI_CLAUDE_FABLE_5_1_CACHE_WRITE_1H,
AZURE_AI_CLAUDE_FABLE_5_1_THINKING,
AZURE_AI_CLAUDE_FABLE_5_1_TOOL_CALL,
AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_INPUT_TEXT,
AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_STREAM,
*(pair.bridged for pair in AZURE_AI_CLAUDE_FABLE_5_1_ROUTE_BRIDGES),
AZURE_AI_CLAUDE_HAIKU_5_5_INPUT_TEXT,
AZURE_AI_CLAUDE_HAIKU_5_5_STREAM,
AZURE_AI_CLAUDE_HAIKU_5_5_CACHE_READ,
AZURE_AI_CLAUDE_HAIKU_5_5_TIERED_INPUT_ABOVE_100K,
AZURE_AI_CLAUDE_HAIKU_5_5_TIERED_CACHE_READ_ABOVE_100K,
AZURE_AI_CLAUDE_HAIKU_5_5_CACHE_WRITE_5M,
AZURE_AI_CLAUDE_HAIKU_5_5_CACHE_WRITE_1H,
AZURE_AI_CLAUDE_HAIKU_5_5_THINKING,
AZURE_AI_CLAUDE_HAIKU_5_5_TOOL_CALL,
AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_INPUT_TEXT,
AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_STREAM,
*(pair.bridged for pair in AZURE_AI_CLAUDE_HAIKU_5_5_ROUTE_BRIDGES),
)
PARITY: Final[tuple[StreamParityTestCase, ...]] = (
StreamParityTestCase(plain=TEXT_COMPLETIONS_OPENAI_BASIC, streamed=TEXT_COMPLETIONS_OPENAI_STREAM_USAGE),
StreamParityTestCase(
plain=AZURE_AI_CLAUDE_FABLE_5_1_INPUT_TEXT, streamed=AZURE_AI_CLAUDE_FABLE_5_1_STREAM
),
StreamParityTestCase(
plain=AZURE_AI_CLAUDE_HAIKU_5_5_INPUT_TEXT, streamed=AZURE_AI_CLAUDE_HAIKU_5_5_STREAM
),
StreamParityTestCase(
plain=AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_INPUT_TEXT,
streamed=AZURE_AI_CLAUDE_FABLE_5_1_MESSAGES_STREAM,
),
StreamParityTestCase(
plain=AZURE_AI_CLAUDE_FABLE_5_1_ROUTE_BRIDGES[0].bridged,
streamed=AZURE_AI_CLAUDE_FABLE_5_1_ROUTE_BRIDGES[1].bridged,
),
StreamParityTestCase(
plain=AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_INPUT_TEXT,
streamed=AZURE_AI_CLAUDE_HAIKU_5_5_MESSAGES_STREAM,
),
StreamParityTestCase(
plain=AZURE_AI_CLAUDE_HAIKU_5_5_ROUTE_BRIDGES[0].bridged,
streamed=AZURE_AI_CLAUDE_HAIKU_5_5_ROUTE_BRIDGES[1].bridged,
),
)
BRIDGE_COST_PARITY: Final[tuple[BridgeCostParityTestCase, ...]] = (
*AZURE_AI_CLAUDE_FABLE_5_1_ROUTE_BRIDGES,
*AZURE_AI_CLAUDE_HAIKU_5_5_ROUTE_BRIDGES,
)
Reference in a new issue View git blame Copy permalink