scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-07 08:26:10 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 89

Merge pull request #39622 from BerriAI/litellm_gpt_6_astra

Browse source 
feat(models): add gpt-6-astra pricing and metadata
This commit is contained in:
Mateo Wang 2026-09-03 13:18:34 -07:00 • committed by GitHub
commit 828d561fcc
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 225 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

69
litellm/model_prices_and_context_window_backup.json
View file

@ -29277,6 +29277,75 @@
"supports_xhigh_reasoning_effort": true,
"supports_minimal_reasoning_effort": true
},
"gpt-6-astra": {
"cache_creation_input_token_cost": 1.25e-05,
"cache_creation_input_token_cost_above_272k_tokens": 2.5e-05,
"cache_creation_input_token_cost_above_272k_tokens_flex": 1.25e-05,
"cache_creation_input_token_cost_above_272k_tokens_priority": 5e-05,
"cache_creation_input_token_cost_flex": 6.25e-06,
"cache_creation_input_token_cost_priority": 2.5e-05,
"cache_read_input_token_cost": 1e-06,
"cache_read_input_token_cost_above_272k_tokens": 2e-06,
"cache_read_input_token_cost_above_272k_tokens_flex": 1e-06,
"cache_read_input_token_cost_above_272k_tokens_priority": 4e-06,
"cache_read_input_token_cost_flex": 5e-07,
"cache_read_input_token_cost_priority": 2e-06,
"input_cost_per_token": 1e-05,
"input_cost_per_token_above_272k_tokens": 2e-05,
"input_cost_per_token_above_272k_tokens_flex": 1e-05,
"input_cost_per_token_above_272k_tokens_priority": 4e-05,
"input_cost_per_token_batches": 5e-06,
"input_cost_per_token_flex": 5e-06,
"input_cost_per_token_priority": 2e-05,
"litellm_provider": "openai",
"max_input_tokens": 922000,
"max_output_tokens": 128000,
"max_tokens": 128000,
"mode": "chat",
"output_cost_per_token": 5e-05,
"output_cost_per_token_above_272k_tokens": 7.5e-05,
"output_cost_per_token_above_272k_tokens_flex": 3.75e-05,
"output_cost_per_token_above_272k_tokens_priority": 0.00015,
"output_cost_per_token_batches": 2.5e-05,
"output_cost_per_token_flex": 2.5e-05,
"output_cost_per_token_priority": 0.0001,
"regional_processing_uplift_multiplier_eu": 1.1,
"regional_processing_uplift_multiplier_us": 1.1,
"search_context_cost_per_query": {
"search_context_size_high": 0.01,
"search_context_size_low": 0.01,
"search_context_size_medium": 0.01
},
"supported_endpoints": [
"/v1/chat/completions",
"/v1/batch",
"/v1/responses"
],
"supported_modalities": [
"text",
"image"
],
"supported_output_modalities": [
"text"
],
"supports_computer_use": true,
"supports_function_calling": true,
"supports_max_reasoning_effort": true,
"supports_minimal_reasoning_effort": false,
"supports_native_streaming": true,
"supports_none_reasoning_effort": false,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_cache_breakpoint": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_web_search": true,
"supports_xhigh_reasoning_effort": true
},
"gpt-5.6": {
"cache_creation_input_token_cost": 5e-06,
"cache_creation_input_token_cost_above_272k_tokens": 1e-05,

69
model_prices_and_context_window.json
View file

@ -29277,6 +29277,75 @@
"supports_xhigh_reasoning_effort": true,
"supports_minimal_reasoning_effort": true
},
"gpt-6-astra": {
"cache_creation_input_token_cost": 1.25e-05,
"cache_creation_input_token_cost_above_272k_tokens": 2.5e-05,
"cache_creation_input_token_cost_above_272k_tokens_flex": 1.25e-05,
"cache_creation_input_token_cost_above_272k_tokens_priority": 5e-05,
"cache_creation_input_token_cost_flex": 6.25e-06,
"cache_creation_input_token_cost_priority": 2.5e-05,
"cache_read_input_token_cost": 1e-06,
"cache_read_input_token_cost_above_272k_tokens": 2e-06,
"cache_read_input_token_cost_above_272k_tokens_flex": 1e-06,
"cache_read_input_token_cost_above_272k_tokens_priority": 4e-06,
"cache_read_input_token_cost_flex": 5e-07,
"cache_read_input_token_cost_priority": 2e-06,
"input_cost_per_token": 1e-05,
"input_cost_per_token_above_272k_tokens": 2e-05,
"input_cost_per_token_above_272k_tokens_flex": 1e-05,
"input_cost_per_token_above_272k_tokens_priority": 4e-05,
"input_cost_per_token_batches": 5e-06,
"input_cost_per_token_flex": 5e-06,
"input_cost_per_token_priority": 2e-05,
"litellm_provider": "openai",
"max_input_tokens": 922000,
"max_output_tokens": 128000,
"max_tokens": 128000,
"mode": "chat",
"output_cost_per_token": 5e-05,
"output_cost_per_token_above_272k_tokens": 7.5e-05,
"output_cost_per_token_above_272k_tokens_flex": 3.75e-05,
"output_cost_per_token_above_272k_tokens_priority": 0.00015,
"output_cost_per_token_batches": 2.5e-05,
"output_cost_per_token_flex": 2.5e-05,
"output_cost_per_token_priority": 0.0001,
"regional_processing_uplift_multiplier_eu": 1.1,
"regional_processing_uplift_multiplier_us": 1.1,
"search_context_cost_per_query": {
"search_context_size_high": 0.01,
"search_context_size_low": 0.01,
"search_context_size_medium": 0.01
},
"supported_endpoints": [
"/v1/chat/completions",
"/v1/batch",
"/v1/responses"
],
"supported_modalities": [
"text",
"image"
],
"supported_output_modalities": [
"text"
],
"supports_computer_use": true,
"supports_function_calling": true,
"supports_max_reasoning_effort": true,
"supports_minimal_reasoning_effort": false,
"supports_native_streaming": true,
"supports_none_reasoning_effort": false,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_cache_breakpoint": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_web_search": true,
"supports_xhigh_reasoning_effort": true
},
"gpt-5.6": {
"cache_creation_input_token_cost": 5e-06,
"cache_creation_input_token_cost_above_272k_tokens": 1e-05,

48
tests/test_litellm/litellm_core_utils/llm_cost_calc/test_llm_cost_calc_utils.py
View file

@ -1662,6 +1662,54 @@ def test_generic_cost_per_token_gpt56_cyber(
assert completion_cost == pytest.approx(completion_tokens * output_rate)
@pytest.mark.parametrize(
"service_tier,tier_multiplier",
[(None, 1.0), ("flex", 0.5), ("priority", 2.0), ("fast", 2.0)],
)
@pytest.mark.parametrize(
"prompt_tokens,input_side_multiplier,output_multiplier",
[(100000, 1.0, 1.0), (300000, 2.0, 1.5)],
)
def test_generic_cost_per_token_gpt_6_astra_price_sheet(
_local_model_cost_map,
service_tier,
tier_multiplier,
prompt_tokens,
input_side_multiplier,
output_multiplier,
):
"""gpt-6-astra launch price sheet: $10 input, $1 cache read, $12.50 cache write, $50 output per 1M tokens.
Above 272K prompt tokens the input-side rates double and the output rate is 1.5x on the whole
request. Flex is half the applicable rate and fast mode, billed as priority, is double it.
"""
cached_tokens = 50000
cache_write_tokens = 40000
text_tokens = prompt_tokens - cached_tokens - cache_write_tokens
completion_tokens = 1000
usage = Usage(
prompt_tokens=prompt_tokens,
completion_tokens=completion_tokens,
total_tokens=prompt_tokens + completion_tokens,
prompt_tokens_details=PromptTokensDetailsWrapper(
cached_tokens=cached_tokens, cache_write_tokens=cache_write_tokens
),
)
prompt_cost, completion_cost = generic_cost_per_token(
model="gpt-6-astra",
usage=usage,
custom_llm_provider="openai",
service_tier=service_tier,
)
input_side = tier_multiplier * input_side_multiplier
assert prompt_cost == pytest.approx(
input_side * (text_tokens * 1e-5 + cached_tokens * 1e-6 + cache_write_tokens * 1.25e-5)
)
assert completion_cost == pytest.approx(tier_multiplier * output_multiplier * completion_tokens * 5e-5)
@pytest.mark.parametrize(
"model,input_cost,output_cost,cache_read_cost",
[

17
tests/test_litellm/router_utils/test_reasoning_effort_capability.py
View file

@ -371,3 +371,20 @@ class TestKimiK3AdvertisesItsDocumentedLevels:
"low",
"high",
)
class TestGpt6AstraAdvertisesItsDocumentedLevels:
def test_the_entry_advertises_low_through_max_without_none(self, local_model_cost_map):
"""OpenAI documents low, medium, high, xhigh and max for gpt-6-astra. Unlike gpt-5.6-sol it
does not take none, so a group must not offer none and must offer max."""
from litellm.utils import _get_model_info_helper
model_info = dict(_get_model_info_helper(model="gpt-6-astra", custom_llm_provider="openai"))
assert resolve_supported_reasoning_efforts(model_info, deployment_is_mapped=True) == (
"low",
"medium",
"high",
"xhigh",
"max",
)

14
tests/test_litellm/test_cost_calculator.py
View file

@ -4473,3 +4473,17 @@ def test_explicit_pricing_precedes_private_provider_response_model(
)
assert selected == expected
def test_batch_cost_calculator_gpt_6_astra_bills_half_the_standard_rate(_local_model_cost_map):
"""gpt-6-astra batch pricing is 50% off the standard $10 input and $50 output rates per 1M tokens."""
from litellm.cost_calculator import batch_cost_calculator
usage = Usage(prompt_tokens=1000, completion_tokens=500, total_tokens=1500)
prompt_cost, completion_cost = batch_cost_calculator(
usage=usage, model="gpt-6-astra", custom_llm_provider="openai"
)
assert prompt_cost == pytest.approx(1000 * 5e-6)
assert completion_cost == pytest.approx(500 * 2.5e-5)

8
tests/test_litellm/test_openai_service_tier_long_context_pricing.py
View file

@ -52,6 +52,12 @@ PRIORITY_LONG_CONTEXT = {
"cache_read_input_token_cost_above_272k_tokens_priority": 8e-08,
"cache_creation_input_token_cost_above_272k_tokens_priority": 1e-06,
},
"gpt-6-astra": {
"input_cost_per_token_above_272k_tokens_priority": 4e-05,
"output_cost_per_token_above_272k_tokens_priority": 0.00015,
"cache_read_input_token_cost_above_272k_tokens_priority": 4e-06,
"cache_creation_input_token_cost_above_272k_tokens_priority": 5e-05,
},
}
EXPECTED = {**FLEX_LONG_CONTEXT, **PRIORITY_LONG_CONTEXT}
@ -63,6 +69,7 @@ NO_PUBLISHED_PRIORITY_LONG_CONTEXT = ("gpt-5.4", "gpt-5.5")
def _local_model_cost_map(monkeypatch: pytest.MonkeyPatch) -> None:
monkeypatch.setenv("LITELLM_LOCAL_MODEL_COST_MAP", "True")
monkeypatch.setattr(litellm, "model_cost", litellm.get_model_cost_map(url=""))
litellm.add_known_models()
@lru_cache(maxsize=2)
@ -114,6 +121,7 @@ TIERED_COST_CASES = [
("gpt-5.6-sol", "priority", 1.6e-05, 6e-05),
("gpt-5.6-terra", "priority", 8e-06, 3.6e-05),
("gpt-5.6-luna", "priority", 8e-07, 3.6e-06),
("gpt-6-astra", "priority", 4e-05, 0.00015),
]