scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-13 23:11:40 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 152

Merge pull request #32025 from BerriAI/litellm_cherrypick_1_90_x

Browse source 
chore(release): backport #31923, #31929, #31393 to stable/1.90.x and cut 1.90.3
This commit is contained in:
yuneng-jiang 2026-07-03 12:17:20 -07:00 • committed by GitHub
commit 2d28d8fadf
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
18 changed files with 942 additions and 237 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
litellm/experimental_mcp_client/client.py
View file

@ -555,7 +555,7 @@ class MCPClient:
Call an MCP Tool.
"""
verbose_logger.info(
f"MCP client calling tool '{call_tool_request_params.name}' with arguments: {call_tool_request_params.arguments}"
f"MCP client calling tool '{call_tool_request_params.name}'"
)
async def on_progress(
@ -664,7 +664,7 @@ class MCPClient:
) -> GetPromptResult:
"""Fetch a prompt definition from the MCP server."""
verbose_logger.info(
f"MCP client fetching prompt '{get_prompt_request_params.name}' with arguments: {get_prompt_request_params.arguments}"
f"MCP client fetching prompt '{get_prompt_request_params.name}'"
)
async def _get_prompt_operation(session: ClientSession):

9
litellm/litellm_core_utils/prompt_templates/factory.py
View file

@ -5547,7 +5547,7 @@ def _bedrock_tools_pt(
]
"""
from litellm.llms.bedrock.common_utils import (
get_bedrock_base_model,
bedrock_converse_supports_strict_tools,
normalize_json_schema_custom_types_to_object,
)
from litellm.litellm_core_utils.prompt_templates.common_utils import unpack_defs
@ -5556,9 +5556,12 @@ def _bedrock_tools_pt(
("array", "boolean", "integer", "null", "number", "object", "string")
)
# Only Claude on Bedrock honours strict tool schemas; other families
# (Nova, Llama, GPT-OSS) reject the strict field outright.
# (Nova, Llama, GPT-OSS) reject the strict field outright. Opus 4.7/4.8
# also reject `strict` on Bedrock Converse (see #31582) — their validator
# maps toolSpec to the native Anthropic tool shape, which has no strict
# field, even though Anthropic's native API accepts it as a top-level key.
supports_strict_tools = bool(
model and get_bedrock_base_model(model).startswith("anthropic")
model and bedrock_converse_supports_strict_tools(model)
)
tool_block_list: List[BedrockToolBlock] = []
for tool_idx, tool in enumerate(tools):

39
litellm/llms/bedrock/chat/converse_transformation.py
View file

@ -76,6 +76,7 @@ from litellm.utils import (
from ..common_utils import (
BedrockError,
BedrockModelInfo,
bedrock_converse_supports_parallel_tool_use_config,
get_anthropic_beta_from_headers,
get_bedrock_tool_name,
is_claude_4_5_on_bedrock,
@ -1176,18 +1177,34 @@ class AmazonConverseConfig(BaseConfig):
if cache_control is None:
return None
cache_point = CachePointBlock(type="default")
if isinstance(cache_control, dict) and "ttl" in cache_control:
ttl = cache_control["ttl"]
if ttl in ["5m", "1h"] and model is not None:
if is_claude_4_5_on_bedrock(model):
cache_point["ttl"] = ttl
cache_point = self._build_cache_point_block(cache_control, model)
if block_type == "system":
return SystemContentBlock(cachePoint=cache_point)
else:
return ContentBlock(cachePoint=cache_point)
@staticmethod
def _build_cache_point_block(
control: Optional[dict], model: Optional[str] = None
) -> CachePointBlock:
"""Build a Bedrock ``cachePoint`` block from an OpenAI-style ``cache_control``/``control`` dict.
``type`` is always ``"default"`` (the only value Bedrock's Converse API
accepts). ``ttl`` is only honored for models that support extended TTL
caching (Claude 4.5 family on Bedrock).
"""
cache_point = CachePointBlock(type="default")
if isinstance(control, dict) and "ttl" in control:
ttl = control["ttl"]
if (
ttl in ["5m", "1h"]
and model is not None
and is_claude_4_5_on_bedrock(model)
):
cache_point["ttl"] = ttl
return cache_point
def _transform_system_message(
self, messages: List[AllMessageValues], model: Optional[str] = None
) -> Tuple[List[AllMessageValues], List[SystemContentBlock]]:
@ -1339,7 +1356,10 @@ class AmazonConverseConfig(BaseConfig):
parallel_tool_use_config = additional_request_params.pop(
"_parallel_tool_use_config", None
)
if parallel_tool_use_config is not None and is_claude_4_5_on_bedrock(model):
if (
parallel_tool_use_config is not None
and bedrock_converse_supports_parallel_tool_use_config(model)
):
for key, value in parallel_tool_use_config.items():
if (
key in additional_request_params
@ -1651,7 +1671,10 @@ class AmazonConverseConfig(BaseConfig):
if cache_injection_points and len(bedrock_tools) > 0:
for point in cache_injection_points:
if point.get("location") == "tool_config":
bedrock_tools.append({"cachePoint": {"type": "default"}})
cache_point = self._build_cache_point_block(
point.get("control"), model
)
bedrock_tools.append(ToolBlock(cachePoint=cache_point))
break
bedrock_tool_config: Optional[ToolConfigBlock] = None

105
litellm/llms/bedrock/common_utils.py
View file

@ -4,9 +4,11 @@ from __future__ import annotations
Common utilities used across bedrock chat/embedding/image generation
"""
import contextlib
import functools
import json
import os
import re
from typing import TYPE_CHECKING, Any, Dict, List, Literal, Optional, Union
if TYPE_CHECKING:
@ -650,39 +652,82 @@ def get_bedrock_base_model(model: str) -> str:
return model
def bedrock_converse_supports_parallel_tool_use_config(model: str) -> bool:
return any(
(litellm.model_cost.get(candidate) or {}).get(
"supports_parallel_tool_use_config"
)
is True
for candidate in (model, get_bedrock_base_model(model))
)
def is_claude_4_5_on_bedrock(model: str) -> bool:
"""
Check if the model is a Claude 4.5 model on Bedrock.
Claude 4.5 models support prompt caching with '5m' and '1h' TTL on Bedrock.
Check if the model supports Bedrock prompt caching with an extended '1h' TTL
(in addition to the default 5m TTL).
Backed by the ``cache_creation_input_token_cost_above_1hr`` field in
``model_prices_and_context_window.json`` instead of a hardcoded list of
model-name patterns, so newly released models pick up support as soon as
their pricing entry ships, with no code change required here.
"""
model_lower = model.lower()
claude_4_5_patterns = [
"sonnet-4.5",
"sonnet_4.5",
"sonnet-4-5",
"sonnet_4_5",
"haiku-4.5",
"haiku_4.5",
"haiku-4-5",
"haiku_4_5",
"opus-4.5",
"opus_4.5",
"opus-4-5",
"opus_4_5",
"sonnet-4.6",
"sonnet_4.6",
"sonnet-4-6",
"sonnet_4_6",
"opus-4.6",
"opus_4.6",
"opus-4-6",
"opus_4_6",
"opus-4.7",
"opus_4.7",
"opus-4-7",
"opus_4_7",
]
return any(pattern in model_lower for pattern in claude_4_5_patterns)
return any(
(litellm.model_cost.get(candidate) or {}).get(
"cache_creation_input_token_cost_above_1hr"
)
is not None
for candidate in (model, get_bedrock_base_model(model))
)
_BEDROCK_MODEL_VERSION_SUFFIX_RE = re.compile(r"-v\d+(?::\d+)?$")
def bedrock_converse_supports_strict_tools(model: str) -> bool:
"""
Whether ``toolSpec.strict`` can be forwarded to Bedrock Converse for ``model``.
Non-Anthropic Bedrock families (Nova, Llama, GPT-OSS) reject the field
outright. Anthropic models forward it unless their entry in
``model_prices_and_context_window.json`` sets
``bedrock_converse_supports_strict_tools: false`` — Bedrock routes those
(Opus 4.7/4.8, see #31582) through a stricter validator that rejects the
``strict`` key on ``toolSpec`` even though Anthropic's native API accepts
it as a top-level tool field.
"""
base = get_bedrock_base_model(model)
if not base.startswith("anthropic"):
return False
flag = _get_bedrock_converse_strict_tools_flag(base)
return flag if flag is not None else True
def _get_bedrock_converse_strict_tools_flag(base_model: str) -> Optional[bool]:
candidates = dict.fromkeys(
(base_model, _BEDROCK_MODEL_VERSION_SUFFIX_RE.sub("", base_model))
)
for candidate in candidates:
with contextlib.suppress(Exception):
model_info = get_cached_model_info()(
model=candidate,
custom_llm_provider="bedrock",
)
flag = model_info.get("bedrock_converse_supports_strict_tools")
if isinstance(flag, bool):
return flag
model_cost_key = model_info.get("key")
if isinstance(model_cost_key, str):
local_flag = (
_get_local_model_cost_map()
.get(model_cost_key, {})
.get("bedrock_converse_supports_strict_tools")
)
if isinstance(local_flag, bool):
return local_flag
return None
def normalize_bedrock_opus_output_config_effort(model: str, output_config: Any) -> None:

173
litellm/model_prices_and_context_window_backup.json
View file

@ -734,7 +734,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-haiku-4-5@20251001": {
"cache_creation_input_token_cost": 1.25e-06,
@ -758,7 +759,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_streaming": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-3-5-sonnet-20240620-v1:0": {
"input_cost_per_token": 3e-06,
@ -777,8 +779,6 @@
"output_cost_per_token_above_200k_tokens": 3e-05,
"cache_creation_input_token_cost_above_200k_tokens": 7.5e-06,
"cache_read_input_token_cost_above_200k_tokens": 6e-07,
"cache_creation_input_token_cost_above_1hr": 7.5e-06,
"cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.5e-05,
"cache_creation_input_token_cost": 3.75e-06,
"cache_read_input_token_cost": 3e-07
},
@ -803,9 +803,7 @@
"input_cost_per_token_above_200k_tokens": 6e-06,
"output_cost_per_token_above_200k_tokens": 3e-05,
"cache_creation_input_token_cost_above_200k_tokens": 7.5e-06,
"cache_read_input_token_cost_above_200k_tokens": 6e-07,
"cache_creation_input_token_cost_above_1hr": 7.5e-06,
"cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.5e-05
"cache_read_input_token_cost_above_200k_tokens": 6e-07
},
"anthropic.claude-3-7-sonnet-20240620-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -981,7 +979,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.25e-06,
@ -1011,7 +1010,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.25e-06,
@ -1041,7 +1041,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.875e-06,
@ -1071,7 +1072,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.875e-06,
@ -1101,7 +1103,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.875e-06,
@ -1131,9 +1134,11 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1163,7 +1168,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-mythos-preview": {
"input_cost_per_token": 0,
@ -1181,6 +1187,7 @@
"supports_output_config": true
},
"global.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1210,9 +1217,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1242,9 +1251,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1274,9 +1285,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1306,7 +1319,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.25e-05,
@ -1339,7 +1353,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.25e-05,
@ -1372,7 +1387,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.375e-05,
@ -1405,7 +1421,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.375e-05,
@ -1438,9 +1455,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1471,9 +1490,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1504,9 +1525,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1537,9 +1560,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1570,9 +1595,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1603,9 +1630,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"jp.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_read_input_token_cost": 5.5e-07,
"input_cost_per_token": 5.5e-06,
@ -1634,7 +1663,8 @@
"tool_use_system_prompt_tokens": 346,
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_minimal_reasoning_effort": true
"supports_minimal_reasoning_effort": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 3.75e-06,
@ -1663,7 +1693,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 3.75e-06,
@ -1692,7 +1723,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1721,7 +1753,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1750,7 +1783,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1779,7 +1813,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"jp.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1808,7 +1843,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-sonnet-4-20250514-v1:0": {
"cache_creation_input_token_cost": 3.75e-06,
@ -1869,7 +1905,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-v1": {
"input_cost_per_token": 8e-06,
@ -2119,7 +2156,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"apac.anthropic.claude-3-sonnet-20240229-v1:0": {
"input_cost_per_token": 3e-06,
@ -2208,7 +2246,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"azure/ada": {
"input_cost_per_token": 1e-07,
@ -9349,7 +9388,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-east-1/claude-sonnet-4-5-20250929-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -9371,7 +9411,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-east-1/meta.llama3-70b-instruct-v1:0": {
"input_cost_per_token": 2.65e-06,
@ -9524,7 +9565,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-west-1/claude-sonnet-4-5-20250929-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -9546,7 +9588,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-west-1/meta.llama3-70b-instruct-v1:0": {
"input_cost_per_token": 2.65e-06,
@ -10276,7 +10319,8 @@
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
"supports_vision": true,
"supports_parallel_tool_use_config": true
},
"claude-opus-4-1": {
"cache_creation_input_token_cost": 1.875e-05,
@ -14076,7 +14120,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-3-5-sonnet-20240620-v1:0": {
"input_cost_per_token": 3e-06,
@ -14289,7 +14334,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"eu.meta.llama3-2-1b-instruct-v1:0": {
"input_cost_per_token": 1.3e-07,
@ -19815,7 +19861,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-sonnet-4-20250514-v1:0": {
"cache_creation_input_token_cost": 3.75e-06,
@ -19867,7 +19914,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"global.amazon.nova-2-lite-v1:0": {
"cache_read_input_token_cost": 7.5e-08,
@ -23976,7 +24024,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"jp.anthropic.claude-haiku-4-5-20251001-v1:0": {
"cache_creation_input_token_cost": 1.375e-06,
@ -23999,7 +24048,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"crusoe/deepseek-ai/DeepSeek-R1-0528": {
"input_cost_per_token": 3e-06,
@ -32544,7 +32594,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-3-5-sonnet-20240620-v1:0": {
"input_cost_per_token": 3e-06,
@ -32703,7 +32754,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"us-gov.anthropic.claude-sonnet-4-5-20250929-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -32730,7 +32782,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-haiku-4-5-20251001-v1:0": {
"cache_creation_input_token_cost": 1.375e-06,
@ -32752,7 +32805,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-20250514-v1:0": {
"cache_creation_input_token_cost": 1.875e-05,
@ -32806,7 +32860,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-opus-4-5-20251101-v1:0": {
"cache_creation_input_token_cost": 6.25e-06,
@ -32835,7 +32890,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-5-20251101-v1:0": {
"cache_creation_input_token_cost": 6.25e-06,
@ -32863,7 +32919,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-sonnet-4-20250514-v1:0": {
"cache_creation_input_token_cost": 3.75e-06,
@ -42775,7 +42832,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_pdf_input": true
"supports_pdf_input": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-west-1/anthropic.claude-haiku-4-5-20251001-v1:0": {
"cache_creation_input_token_cost": 1.5e-06,
@ -42798,7 +42856,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_pdf_input": true
"supports_pdf_input": true,
"supports_parallel_tool_use_config": true
},
"soniox/stt-async-v4": {
"litellm_provider": "soniox",

1
litellm/types/integrations/anthropic_cache_control_hook.py
View file

@ -20,6 +20,7 @@ class CacheControlToolConfigInjectionPoint(TypedDict):
"""Type for tool_config-level injection points (Bedrock)."""
location: Literal["tool_config"]
control: Optional[ChatCompletionCachedContent]
CacheControlInjectionPoint = Union[

1
litellm/types/utils.py
View file

@ -151,6 +151,7 @@ class ProviderSpecificModelInfo(TypedDict, total=False):
bedrock_output_config_effort_ceiling: Optional[
Literal["low", "medium", "high", "max", "xhigh"]
]
bedrock_converse_supports_strict_tools: Optional[bool]
class SearchContextCostPerQuery(TypedDict, total=False):

3
litellm/utils.py
View file

@ -6221,6 +6221,9 @@ def _get_model_info_helper(
bedrock_output_config_effort_ceiling=_model_info.get(
"bedrock_output_config_effort_ceiling", None
),
bedrock_converse_supports_strict_tools=_model_info.get(
"bedrock_converse_supports_strict_tools", None
),
supports_computer_use=_model_info.get("supports_computer_use", None),
search_context_cost_per_query=_model_info.get(
"search_context_cost_per_query", None

173
model_prices_and_context_window.json
View file

@ -734,7 +734,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-haiku-4-5@20251001": {
"cache_creation_input_token_cost": 1.25e-06,
@ -758,7 +759,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_streaming": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-3-5-sonnet-20240620-v1:0": {
"input_cost_per_token": 3e-06,
@ -777,8 +779,6 @@
"output_cost_per_token_above_200k_tokens": 3e-05,
"cache_creation_input_token_cost_above_200k_tokens": 7.5e-06,
"cache_read_input_token_cost_above_200k_tokens": 6e-07,
"cache_creation_input_token_cost_above_1hr": 7.5e-06,
"cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.5e-05,
"cache_creation_input_token_cost": 3.75e-06,
"cache_read_input_token_cost": 3e-07
},
@ -803,9 +803,7 @@
"input_cost_per_token_above_200k_tokens": 6e-06,
"output_cost_per_token_above_200k_tokens": 3e-05,
"cache_creation_input_token_cost_above_200k_tokens": 7.5e-06,
"cache_read_input_token_cost_above_200k_tokens": 6e-07,
"cache_creation_input_token_cost_above_1hr": 7.5e-06,
"cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.5e-05
"cache_read_input_token_cost_above_200k_tokens": 6e-07
},
"anthropic.claude-3-7-sonnet-20240620-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -981,7 +979,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.25e-06,
@ -1011,7 +1010,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.25e-06,
@ -1041,7 +1041,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.875e-06,
@ -1071,7 +1072,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.875e-06,
@ -1101,7 +1103,8 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-opus-4-6-v1": {
"cache_creation_input_token_cost": 6.875e-06,
@ -1131,9 +1134,11 @@
"supports_native_structured_output": true,
"supports_output_config": true,
"supports_max_reasoning_effort": true,
"bedrock_output_config_effort_ceiling": "max"
"bedrock_output_config_effort_ceiling": "max",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1163,7 +1168,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-mythos-preview": {
"input_cost_per_token": 0,
@ -1181,6 +1187,7 @@
"supports_output_config": true
},
"global.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1210,9 +1217,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1242,9 +1251,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1274,9 +1285,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1306,7 +1319,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.25e-05,
@ -1339,7 +1353,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.25e-05,
@ -1372,7 +1387,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.375e-05,
@ -1405,7 +1421,8 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-fable-5": {
"cache_creation_input_token_cost": 1.375e-05,
@ -1438,9 +1455,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1471,9 +1490,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.25e-06,
"cache_creation_input_token_cost_above_1hr": 1e-05,
"cache_read_input_token_cost": 5e-07,
@ -1504,9 +1525,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1537,9 +1560,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1570,9 +1595,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-opus-4-8": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_creation_input_token_cost_above_1hr": 1.1e-05,
"cache_read_input_token_cost": 5.5e-07,
@ -1603,9 +1630,11 @@
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "xhigh"
"bedrock_output_config_effort_ceiling": "xhigh",
"supports_parallel_tool_use_config": true
},
"jp.anthropic.claude-opus-4-7": {
"bedrock_converse_supports_strict_tools": false,
"cache_creation_input_token_cost": 6.875e-06,
"cache_read_input_token_cost": 5.5e-07,
"input_cost_per_token": 5.5e-06,
@ -1634,7 +1663,8 @@
"tool_use_system_prompt_tokens": 346,
"supports_native_structured_output": true,
"supports_max_reasoning_effort": true,
"supports_minimal_reasoning_effort": true
"supports_minimal_reasoning_effort": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 3.75e-06,
@ -1663,7 +1693,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 3.75e-06,
@ -1692,7 +1723,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1721,7 +1753,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1750,7 +1783,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1779,7 +1813,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"jp.anthropic.claude-sonnet-4-6": {
"cache_creation_input_token_cost": 4.125e-06,
@ -1808,7 +1843,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true
"supports_output_config": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-sonnet-4-20250514-v1:0": {
"cache_creation_input_token_cost": 3.75e-06,
@ -1869,7 +1905,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"anthropic.claude-v1": {
"input_cost_per_token": 8e-06,
@ -2119,7 +2156,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"apac.anthropic.claude-3-sonnet-20240229-v1:0": {
"input_cost_per_token": 3e-06,
@ -2208,7 +2246,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"azure/ada": {
"input_cost_per_token": 1e-07,
@ -9349,7 +9388,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-east-1/claude-sonnet-4-5-20250929-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -9371,7 +9411,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-east-1/meta.llama3-70b-instruct-v1:0": {
"input_cost_per_token": 2.65e-06,
@ -9524,7 +9565,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-west-1/claude-sonnet-4-5-20250929-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -9546,7 +9588,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-west-1/meta.llama3-70b-instruct-v1:0": {
"input_cost_per_token": 2.65e-06,
@ -10276,7 +10319,8 @@
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
"supports_vision": true,
"supports_parallel_tool_use_config": true
},
"claude-opus-4-1": {
"cache_creation_input_token_cost": 1.875e-05,
@ -14084,7 +14128,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-3-5-sonnet-20240620-v1:0": {
"input_cost_per_token": 3e-06,
@ -14297,7 +14342,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"eu.meta.llama3-2-1b-instruct-v1:0": {
"input_cost_per_token": 1.3e-07,
@ -19823,7 +19869,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-sonnet-4-20250514-v1:0": {
"cache_creation_input_token_cost": 3.75e-06,
@ -19875,7 +19922,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"global.amazon.nova-2-lite-v1:0": {
"cache_read_input_token_cost": 7.5e-08,
@ -23984,7 +24032,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"jp.anthropic.claude-haiku-4-5-20251001-v1:0": {
"cache_creation_input_token_cost": 1.375e-06,
@ -24007,7 +24056,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"crusoe/deepseek-ai/DeepSeek-R1-0528": {
"input_cost_per_token": 3e-06,
@ -32592,7 +32642,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-3-5-sonnet-20240620-v1:0": {
"input_cost_per_token": 3e-06,
@ -32751,7 +32802,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"us-gov.anthropic.claude-sonnet-4-5-20250929-v1:0": {
"cache_creation_input_token_cost": 4.5e-06,
@ -32778,7 +32830,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"au.anthropic.claude-haiku-4-5-20251001-v1:0": {
"cache_creation_input_token_cost": 1.375e-06,
@ -32800,7 +32853,8 @@
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true
"supports_native_structured_output": true,
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-opus-4-20250514-v1:0": {
"cache_creation_input_token_cost": 1.875e-05,
@ -32854,7 +32908,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"global.anthropic.claude-opus-4-5-20251101-v1:0": {
"cache_creation_input_token_cost": 6.25e-06,
@ -32883,7 +32938,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"eu.anthropic.claude-opus-4-5-20251101-v1:0": {
"cache_creation_input_token_cost": 6.25e-06,
@ -32911,7 +32967,8 @@
"supports_vision": true,
"supports_native_structured_output": true,
"supports_output_config": true,
"bedrock_output_config_effort_ceiling": "high"
"bedrock_output_config_effort_ceiling": "high",
"supports_parallel_tool_use_config": true
},
"us.anthropic.claude-sonnet-4-20250514-v1:0": {
"cache_creation_input_token_cost": 3.75e-06,
@ -42985,7 +43042,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_pdf_input": true
"supports_pdf_input": true,
"supports_parallel_tool_use_config": true
},
"bedrock/us-gov-west-1/anthropic.claude-haiku-4-5-20251001-v1:0": {
"cache_creation_input_token_cost": 1.5e-06,
@ -43008,7 +43066,8 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_native_structured_output": true,
"supports_pdf_input": true
"supports_pdf_input": true,
"supports_parallel_tool_use_config": true
},
"snowflake/claude-sonnet-4-5": {
"max_tokens": 16384,

4
pyproject.toml
View file

@ -1,6 +1,6 @@
[project]
name = "litellm"
version = "1.90.2"
version = "1.90.3"
description = "Library to easily interface with LLM API providers"
readme = "README.md"
requires-python = ">=3.10, <3.14"
@ -272,7 +272,7 @@ source-exclude = [
profile = "black"
[tool.commitizen]
version = "1.90.2"
version = "1.90.3"
version_files = [
"pyproject.toml:^version",
]

46
tests/test_litellm/experimental_mcp_client/test_mcp_client.py
View file

@ -543,5 +543,51 @@ class TestExecuteSessionOperationSurfacesTransportError:
assert result == "done"
def _all_logged_messages(mock_logger):
return " ".join(
str(call.args[0])
for level in ("info", "debug", "warning", "error", "exception")
for call in getattr(mock_logger, level).call_args_list
if call.args
)
@pytest.mark.asyncio
async def test_call_tool_does_not_log_arguments():
from mcp.types import CallToolRequestParams
secret = "ssn-123-45-6789"
client = MCPClient(server_url="http://test-server")
client.run_with_session = AsyncMock(return_value=MagicMock())
params = CallToolRequestParams(
name="search_tool", arguments={"input": secret, "model": "gpt-5-mini"}
)
with patch.object(mcp_client_module, "verbose_logger") as mock_logger:
await client.call_tool(params)
logged = _all_logged_messages(mock_logger)
assert "search_tool" in logged
assert secret not in logged
assert "gpt-5-mini" not in logged
@pytest.mark.asyncio
async def test_get_prompt_does_not_log_arguments():
from mcp.types import GetPromptRequestParams
secret = "ssn-987-65-4321"
client = MCPClient(server_url="http://test-server")
client.run_with_session = AsyncMock(return_value=MagicMock())
params = GetPromptRequestParams(name="my_prompt", arguments={"input": secret})
with patch.object(mcp_client_module, "verbose_logger") as mock_logger:
await client.get_prompt(params)
logged = _all_logged_messages(mock_logger)
assert "my_prompt" in logged
assert secret not in logged
if __name__ == "__main__":
pytest.main([__file__])

124
tests/test_litellm/litellm_core_utils/prompt_templates/test_bedrock_converse_strict_tools_opus_47_48.py Normal file
View file

@ -0,0 +1,124 @@
"""Regression tests for Bedrock Converse ``toolSpec.strict`` forwarding.
Bedrock Converse routes Claude Opus 4.7/4.8 through an Anthropic-compatible
validator that rejects ``toolSpec.strict`` even though Anthropic's native API
accepts ``strict`` as a top-level tool field for the same models. See
BerriAI/litellm#31582.
"""
import pytest
from litellm.litellm_core_utils.prompt_templates.factory import _bedrock_tools_pt
from litellm.llms.bedrock.common_utils import bedrock_converse_supports_strict_tools
_STRICT_TOOL = [
{
"type": "function",
"function": {
"name": "get_weather",
"strict": True,
"description": "Get the weather for a city",
"parameters": {
"type": "object",
"properties": {
"city": {"type": "string"},
"unit": {"type": "string", "enum": ["celsius"]},
},
"required": ["city", "unit"],
"additionalProperties": False,
},
},
}
]
@pytest.mark.parametrize(
"model_id",
[
"bedrock/us.anthropic.claude-opus-4-7",
"bedrock/us.anthropic.claude-opus-4-8",
"anthropic.claude-opus-4-7",
"anthropic.claude-opus-4-8",
"anthropic.claude-opus-4-7-v1:0",
"bedrock/eu.anthropic.claude-opus-4-8-v1:0",
"bedrock/global.anthropic.claude-opus-4-7",
],
)
def test_bedrock_tools_pt_strict_dropped_for_opus_47_48(model_id: str) -> None:
"""Opus 4.7/4.8 on Bedrock Converse reject toolSpec.strict — must be dropped."""
result = _bedrock_tools_pt(_STRICT_TOOL, model=model_id)
assert (
"strict" not in result[0]["toolSpec"]
), f"strict leaked into toolSpec for {model_id}: {result[0]['toolSpec']}"
@pytest.mark.parametrize(
"model_id",
[
"anthropic.claude-sonnet-4-5-20250929-v1:0",
"bedrock/us.anthropic.claude-sonnet-4-6",
"bedrock/us.anthropic.claude-opus-4-6",
"bedrock/us.anthropic.claude-opus-4-5",
],
)
def test_bedrock_tools_pt_strict_kept_for_other_anthropic(model_id: str) -> None:
"""Sonnet 4.5/4.6 and Opus <=4.6 accept toolSpec.strict — keep forwarding it."""
result = _bedrock_tools_pt(_STRICT_TOOL, model=model_id)
assert (
result[0]["toolSpec"]["strict"] is True
), f"strict missing for {model_id}: {result[0]['toolSpec']}"
@pytest.mark.parametrize(
"model_id",
[
"us.amazon.nova-micro-v1:0",
"meta.llama3-2-11b-instruct-v1:0",
],
)
def test_bedrock_tools_pt_strict_dropped_for_non_anthropic(model_id: str) -> None:
"""Non-Anthropic Bedrock families reject toolSpec.strict — must be dropped."""
result = _bedrock_tools_pt(_STRICT_TOOL, model=model_id)
assert "strict" not in result[0]["toolSpec"]
def test_bedrock_converse_supports_strict_tools_helper() -> None:
"""Direct check for the gate helper used by factory.py."""
assert (
bedrock_converse_supports_strict_tools("bedrock/us.anthropic.claude-opus-4-7")
is False
)
assert (
bedrock_converse_supports_strict_tools("bedrock/us.anthropic.claude-opus-4-8")
is False
)
assert (
bedrock_converse_supports_strict_tools(
"anthropic.claude-sonnet-4-5-20250929-v1:0"
)
is True
)
assert (
bedrock_converse_supports_strict_tools("bedrock/us.anthropic.claude-opus-4-6")
is True
)
assert bedrock_converse_supports_strict_tools("us.amazon.nova-micro-v1:0") is False
assert bedrock_converse_supports_strict_tools("") is False
@pytest.mark.parametrize(
"cost_map_key",
[
"anthropic.claude-opus-4-7",
"us.anthropic.claude-opus-4-7",
"anthropic.claude-opus-4-8",
"us.anthropic.claude-opus-4-8",
],
)
def test_strict_tools_flag_set_in_model_cost_map(cost_map_key: str) -> None:
"""The gate is driven by ``bedrock_converse_supports_strict_tools: false`` in
``model_prices_and_context_window.json``, not hardcoded model patterns."""
from litellm.litellm_core_utils.get_model_cost_map import GetModelCostMap
cost_map = GetModelCostMap.load_local_model_cost_map()
assert cost_map[cost_map_key]["bedrock_converse_supports_strict_tools"] is False

177
tests/test_litellm/litellm_core_utils/prompt_templates/test_litellm_core_utils_prompt_templates_factory.py
View file

@ -1,5 +1,6 @@
import base64
import json
import os
from unittest.mock import MagicMock, patch
import pytest
@ -2608,102 +2609,132 @@ def test_add_cache_point_tool_block_passes_ttl_for_claude_4_5():
TTL ordering constraint (tools -> system -> messages).
Ref: https://github.com/BerriAI/litellm/issues/XXXXX
Forces the bundled local cost map so ttl eligibility (driven by
`cache_creation_input_token_cost_above_1hr` in litellm.model_cost) reads
this branch's pricing data rather than the network-fetched `main` copy,
which lacks the fix until merge.
"""
from litellm.litellm_core_utils.prompt_templates.factory import (
add_cache_point_tool_block,
)
tool_with_1h = {
"type": "function",
"function": {"name": "get_weather", "parameters": {"type": "object"}},
"cache_control": {"type": "ephemeral", "ttl": "1h"},
}
old_env = os.environ.get("LITELLM_LOCAL_MODEL_COST_MAP")
old_cost = litellm.model_cost
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
try:
tool_with_1h = {
"type": "function",
"function": {"name": "get_weather", "parameters": {"type": "object"}},
"cache_control": {"type": "ephemeral", "ttl": "1h"},
}
# Claude 4.5 model: ttl should be preserved
result = add_cache_point_tool_block(
tool_with_1h, model="us.anthropic.claude-sonnet-4-5-20250514-v1:0"
)
assert result is not None
assert result["cachePoint"]["type"] == "default"
assert result["cachePoint"]["ttl"] == "1h"
# Claude 4.5 model: ttl should be preserved
result = add_cache_point_tool_block(
tool_with_1h, model="jp.anthropic.claude-opus-4-7"
)
assert result is not None
assert result["cachePoint"]["type"] == "default"
assert result["cachePoint"]["ttl"] == "1h"
# Claude 4.5 model with 5m ttl: also preserved
tool_with_5m = {
"cache_control": {"type": "ephemeral", "ttl": "5m"},
}
result_5m = add_cache_point_tool_block(
tool_with_5m, model="us.anthropic.claude-sonnet-4-5-20250514-v1:0"
)
assert result_5m is not None
assert result_5m["cachePoint"]["ttl"] == "5m"
# Claude 4.5 model with 5m ttl: also preserved
tool_with_5m = {
"cache_control": {"type": "ephemeral", "ttl": "5m"},
}
result_5m = add_cache_point_tool_block(
tool_with_5m, model="jp.anthropic.claude-opus-4-7"
)
assert result_5m is not None
assert result_5m["cachePoint"]["ttl"] == "5m"
# Older model: ttl should be stripped
result_old = add_cache_point_tool_block(
tool_with_1h, model="anthropic.claude-3-5-sonnet-20241022-v2:0"
)
assert result_old is not None
assert result_old["cachePoint"]["type"] == "default"
assert "ttl" not in result_old["cachePoint"]
# Older model: ttl should be stripped
result_old = add_cache_point_tool_block(
tool_with_1h, model="anthropic.claude-3-5-sonnet-20241022-v2:0"
)
assert result_old is not None
assert result_old["cachePoint"]["type"] == "default"
assert "ttl" not in result_old["cachePoint"]
# No model provided: ttl should be stripped (safe default)
result_no_model = add_cache_point_tool_block(tool_with_1h, model=None)
assert result_no_model is not None
assert "ttl" not in result_no_model["cachePoint"]
# No model provided: ttl should be stripped (safe default)
result_no_model = add_cache_point_tool_block(tool_with_1h, model=None)
assert result_no_model is not None
assert "ttl" not in result_no_model["cachePoint"]
# No cache_control: returns None (unchanged behavior)
tool_no_cache = {
"type": "function",
"function": {"name": "get_weather", "parameters": {"type": "object"}},
}
assert add_cache_point_tool_block(tool_no_cache) is None
# No cache_control: returns None (unchanged behavior)
tool_no_cache = {
"type": "function",
"function": {"name": "get_weather", "parameters": {"type": "object"}},
}
assert add_cache_point_tool_block(tool_no_cache) is None
# cache_control without ttl: returns default cachePoint (unchanged behavior)
tool_no_ttl = {"cache_control": {"type": "ephemeral"}}
result_no_ttl = add_cache_point_tool_block(
tool_no_ttl, model="us.anthropic.claude-sonnet-4-5-20250514-v1:0"
)
assert result_no_ttl is not None
assert result_no_ttl["cachePoint"]["type"] == "default"
assert "ttl" not in result_no_ttl["cachePoint"]
# cache_control without ttl: returns default cachePoint (unchanged behavior)
tool_no_ttl = {"cache_control": {"type": "ephemeral"}}
result_no_ttl = add_cache_point_tool_block(
tool_no_ttl, model="us.anthropic.claude-sonnet-4-5-20250929-v1:0"
)
assert result_no_ttl is not None
assert result_no_ttl["cachePoint"]["type"] == "default"
assert "ttl" not in result_no_ttl["cachePoint"]
finally:
litellm.model_cost = old_cost
if old_env is None:
os.environ.pop("LITELLM_LOCAL_MODEL_COST_MAP", None)
else:
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = old_env
def test_bedrock_tools_pt_passes_ttl_for_claude_4_5():
"""
End-to-end: _bedrock_tools_pt should produce cachePoint blocks with ttl
for Claude 4.5+ models when tools have cache_control with ttl.
Forces the bundled local cost map so ttl eligibility (driven by
`cache_creation_input_token_cost_above_1hr` in litellm.model_cost) reads
this branch's pricing data rather than the network-fetched `main` copy,
which lacks the fix until merge.
"""
from litellm.litellm_core_utils.prompt_templates.factory import _bedrock_tools_pt
tools = [
{
"type": "function",
"function": {
"name": "get_weather",
"description": "Get weather",
"parameters": {
"type": "object",
"properties": {"city": {"type": "string"}},
old_env = os.environ.get("LITELLM_LOCAL_MODEL_COST_MAP")
old_cost = litellm.model_cost
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
try:
tools = [
{
"type": "function",
"function": {
"name": "get_weather",
"description": "Get weather",
"parameters": {
"type": "object",
"properties": {"city": {"type": "string"}},
},
},
},
"cache_control": {"type": "ephemeral", "ttl": "1h"},
}
]
"cache_control": {"type": "ephemeral", "ttl": "1h"},
}
]
# Claude 4.5: cachePoint should have ttl
result = _bedrock_tools_pt(
tools, model="us.anthropic.claude-sonnet-4-5-20250514-v1:0"
)
cache_blocks = [b for b in result if "cachePoint" in b]
assert len(cache_blocks) == 1
assert cache_blocks[0]["cachePoint"]["ttl"] == "1h"
# Claude 4.5: cachePoint should have ttl
result = _bedrock_tools_pt(tools, model="jp.anthropic.claude-opus-4-7")
cache_blocks = [b for b in result if "cachePoint" in b]
assert len(cache_blocks) == 1
assert cache_blocks[0]["cachePoint"]["ttl"] == "1h"
# Older model: cachePoint should not have ttl
result_old = _bedrock_tools_pt(
tools, model="anthropic.claude-3-5-sonnet-20241022-v2:0"
)
cache_blocks_old = [b for b in result_old if "cachePoint" in b]
assert len(cache_blocks_old) == 1
assert "ttl" not in cache_blocks_old[0]["cachePoint"]
# Older model: cachePoint should not have ttl
result_old = _bedrock_tools_pt(
tools, model="anthropic.claude-3-5-sonnet-20241022-v2:0"
)
cache_blocks_old = [b for b in result_old if "cachePoint" in b]
assert len(cache_blocks_old) == 1
assert "ttl" not in cache_blocks_old[0]["cachePoint"]
finally:
litellm.model_cost = old_cost
if old_env is None:
os.environ.pop("LITELLM_LOCAL_MODEL_COST_MAP", None)
else:
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = old_env
def test_convert_to_anthropic_tool_result_openai_file_pdf_becomes_document():

261
tests/test_litellm/llms/bedrock/chat/test_converse_transformation.py
View file

@ -611,6 +611,65 @@ def test_transform_request_helper_includes_anthropic_beta_and_tools():
assert fields["tools"][0]["type"] == "computer_20250124"
def test_parallel_tool_calls_config_kept_for_flagged_model():
old_env = os.environ.get("LITELLM_LOCAL_MODEL_COST_MAP")
old_cost = litellm.model_cost
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
try:
config = AmazonConverseConfig()
optional_params = config.map_openai_params(
model="anthropic.claude-sonnet-4-6",
non_default_params={"parallel_tool_calls": False},
optional_params={},
drop_params=False,
)
data = config._transform_request_helper(
model="anthropic.claude-sonnet-4-6",
system_content_blocks=[],
optional_params=optional_params,
messages=None,
)
assert data["additionalModelRequestFields"]["tool_choice"] == {
"disable_parallel_tool_use": True
}
finally:
litellm.model_cost = old_cost
if old_env is None:
os.environ.pop("LITELLM_LOCAL_MODEL_COST_MAP", None)
else:
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = old_env
def test_parallel_tool_calls_config_dropped_for_ttl_only_model(
monkeypatch: pytest.MonkeyPatch,
):
model = "anthropic.claude-fable-5"
monkeypatch.setitem(
litellm.model_cost,
model,
{"cache_creation_input_token_cost_above_1hr": 2e-05},
)
config = AmazonConverseConfig()
optional_params = config.map_openai_params(
model=model,
non_default_params={"parallel_tool_calls": False},
optional_params={},
drop_params=False,
)
data = config._transform_request_helper(
model=model,
system_content_blocks=[],
optional_params=optional_params,
messages=None,
)
assert "tool_choice" not in data.get("additionalModelRequestFields", {})
def test_transform_response_with_computer_use_tool():
"""Test response transformation with computer use tool call."""
import httpx
@ -4098,8 +4157,10 @@ _TOOL_PARAM = [
]
def test_parallel_tool_calls_newer_model_adds_disable_flag():
def test_parallel_tool_calls_newer_model_adds_disable_flag(monkeypatch):
"""Newer Claude models (4.5+) should get disable_parallel_tool_use in additionalModelRequestFields."""
monkeypatch.setenv("LITELLM_LOCAL_MODEL_COST_MAP", "True")
monkeypatch.setattr(litellm, "model_cost", litellm.get_model_cost_map(url=""))
config = AmazonConverseConfig()
model = "anthropic.claude-sonnet-4-5-20250929-v1:0"
messages = [{"role": "user", "content": "What's the weather in SF and NYC?"}]
@ -4130,6 +4191,44 @@ def test_parallel_tool_calls_newer_model_adds_disable_flag():
assert "parallel_tool_calls" not in request_data["additionalModelRequestFields"]
def test_parallel_tool_calls_flag_decoupled_from_ttl_pricing(monkeypatch):
"""
The disable_parallel_tool_use gate must read supports_parallel_tool_use_config,
not the 1h-TTL pricing field: a model carrying only the former still gets the flag.
"""
from litellm.llms.bedrock.common_utils import is_claude_4_5_on_bedrock
config = AmazonConverseConfig()
model = "anthropic.claude-parallel-tool-use-only"
monkeypatch.setitem(
litellm.model_cost, model, {"supports_parallel_tool_use_config": True}
)
assert is_claude_4_5_on_bedrock(model) is False
messages = [{"role": "user", "content": "What's the weather in SF and NYC?"}]
optional_params = config.map_openai_params(
non_default_params={"parallel_tool_calls": False, "tools": _TOOL_PARAM},
optional_params={},
model=model,
drop_params=False,
)
request_data = config.transform_request(
model=model,
messages=messages,
optional_params=optional_params,
litellm_params={},
headers={},
)
assert (
request_data["additionalModelRequestFields"]["tool_choice"][
"disable_parallel_tool_use"
]
is True
)
def test_parallel_tool_calls_older_model_drops_disable_flag():
"""Older Claude models (pre-4.5) must NOT receive disable_parallel_tool_use — Bedrock rejects it."""
config = AmazonConverseConfig()
@ -4554,6 +4653,166 @@ def test_cache_control_injection_tool_config_not_added_without_injection_point()
assert all("cachePoint" not in tool for tool in tools)
def test_cache_control_injection_tool_config_honors_ttl_for_supported_model():
"""
Regression test: cache_control_injection_points with location=tool_config
must honor the requested `control.ttl`, mirroring the message/system
cache_control behavior, instead of always emitting a bare
{"type": "default"} cachePoint with no ttl.
Forces the bundled local cost map so `is_claude_4_5_on_bedrock` (which
reads `cache_creation_input_token_cost_above_1hr` from litellm.model_cost)
sees this branch's pricing data rather than the network-fetched `main`
copy, which lacks it until merge.
"""
old_env = os.environ.get("LITELLM_LOCAL_MODEL_COST_MAP")
old_cost = litellm.model_cost
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
try:
config = AmazonConverseConfig()
messages = [
{"role": "user", "content": "What is the weather?"},
]
optional_params = {
"tools": [
{
"type": "function",
"function": {
"name": "get_weather",
"description": "Get weather for a location",
"parameters": {
"type": "object",
"properties": {"location": {"type": "string"}},
"required": ["location"],
},
},
}
],
"cache_control_injection_points": [
{
"location": "tool_config",
"control": {"type": "ephemeral", "ttl": "1h"},
},
],
}
result = config._transform_request(
model="us.anthropic.claude-sonnet-4-5-20250929-v1:0",
messages=messages,
optional_params=optional_params,
litellm_params={},
)
tools = result["toolConfig"]["tools"]
assert tools[-1] == {"cachePoint": {"type": "default", "ttl": "1h"}}
finally:
litellm.model_cost = old_cost
if old_env is None:
os.environ.pop("LITELLM_LOCAL_MODEL_COST_MAP", None)
else:
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = old_env
def test_cache_control_injection_tool_config_honors_ttl_for_regional_model_lacking_own_pricing():
"""
Regression test: a regional pricing entry that omits
`cache_creation_input_token_cost_above_1hr` (e.g. `jp.anthropic.claude-opus-4-7`)
must not shadow the base model entry that carries it; the requested ttl
survives through the base-model fallback.
"""
old_env = os.environ.get("LITELLM_LOCAL_MODEL_COST_MAP")
old_cost = litellm.model_cost
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
try:
assert (
"cache_creation_input_token_cost_above_1hr"
not in litellm.model_cost["jp.anthropic.claude-opus-4-7"]
)
assert (
"cache_creation_input_token_cost_above_1hr"
in litellm.model_cost["anthropic.claude-opus-4-7"]
)
config = AmazonConverseConfig()
messages = [
{"role": "user", "content": "What is the weather?"},
]
optional_params = {
"tools": [
{
"type": "function",
"function": {
"name": "get_weather",
"description": "Get weather for a location",
"parameters": {
"type": "object",
"properties": {"location": {"type": "string"}},
"required": ["location"],
},
},
}
],
"cache_control_injection_points": [
{
"location": "tool_config",
"control": {"type": "ephemeral", "ttl": "1h"},
},
],
}
result = config._transform_request(
model="jp.anthropic.claude-opus-4-7",
messages=messages,
optional_params=optional_params,
litellm_params={},
)
tools = result["toolConfig"]["tools"]
assert tools[-1] == {"cachePoint": {"type": "default", "ttl": "1h"}}
finally:
litellm.model_cost = old_cost
if old_env is None:
os.environ.pop("LITELLM_LOCAL_MODEL_COST_MAP", None)
else:
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = old_env
def test_cache_control_injection_tool_config_drops_ttl_for_unsupported_model():
"""
Models that don't support extended TTL caching (only Claude 4.5+ on
Bedrock does) must fall back to the default cachePoint with no ttl,
even if the caller requested one, matching message/system behavior.
"""
config = AmazonConverseConfig()
messages = [
{"role": "user", "content": "What is the weather?"},
]
optional_params = {
"tools": [
{
"type": "function",
"function": {
"name": "get_weather",
"description": "Get weather for a location",
"parameters": {
"type": "object",
"properties": {"location": {"type": "string"}},
"required": ["location"],
},
},
}
],
"cache_control_injection_points": [
{"location": "tool_config", "control": {"type": "ephemeral", "ttl": "1h"}},
],
}
result = config._transform_request(
model="anthropic.claude-3-5-haiku-20241022-v1:0",
messages=messages,
optional_params=optional_params,
litellm_params={},
)
tools = result["toolConfig"]["tools"]
assert tools[-1] == {"cachePoint": {"type": "default"}}
def test_translate_response_format_json_schema_still_injects_tool():
"""
response_format with an explicit json_schema should still use the

25
tests/test_litellm/llms/bedrock/messages/invoke_transformations/test_anthropic_claude3_transformation.py
View file

@ -25,6 +25,23 @@ from litellm.llms.bedrock.messages.invoke_transformations.anthropic_claude3_tran
)
@pytest.fixture
def local_model_cost_map(monkeypatch):
"""Force the bundled backup cost map so capability detection reads this
branch's flags, which the network-fetched ``main`` copy may lack."""
import litellm
original = litellm.model_cost
monkeypatch.setenv("LITELLM_LOCAL_MODEL_COST_MAP", "True")
litellm.model_cost = litellm.get_model_cost_map(url="")
litellm.get_model_info.cache_clear()
try:
yield
finally:
litellm.model_cost = original
litellm.get_model_info.cache_clear()
@pytest.mark.asyncio
async def test_bedrock_sse_wrapper_encodes_dict_chunks():
"""Verify that `bedrock_sse_wrapper` converts dictionary chunks to properly formatted Server-Sent Events and forwards non-dict chunks unchanged."""
@ -467,7 +484,7 @@ def test_bedrock_invoke_messages_transform_converts_custom_tool_schema_type_to_o
assert result["tools"][0]["type"] == "custom"
def test_remove_ttl_from_cache_control_processes_tools():
def test_remove_ttl_from_cache_control_processes_tools(local_model_cost_map):
"""
Ensure _remove_ttl_from_cache_control also sanitizes cache_control on tools.
@ -513,7 +530,9 @@ def test_remove_ttl_from_cache_control_processes_tools():
assert "ttl" not in request["system"][0]["cache_control"]
def test_remove_ttl_from_cache_control_preserves_tools_ttl_for_claude_4_5():
def test_remove_ttl_from_cache_control_preserves_tools_ttl_for_claude_4_5(
local_model_cost_map,
):
"""
For Claude 4.5+ models, ttl in ["5m", "1h"] should be preserved on tools,
just like it is for system and messages.
@ -539,7 +558,7 @@ def test_remove_ttl_from_cache_control_preserves_tools_ttl_for_claude_4_5():
}
cfg._remove_ttl_from_cache_control(
request, model="us.anthropic.claude-sonnet-4-5-20250514-v1:0"
request, model="us.anthropic.claude-sonnet-4-5-20250929-v1:0"
)
# Both tools and system should preserve ttl for Claude 4.5

30
tests/test_litellm/llms/bedrock/test_bedrock_common_utils.py
View file

@ -263,3 +263,33 @@ def test_bundled_bedrock_opus_model_info_declares_output_config_effort_ceiling(
model_info = GetModelCostMap.load_local_model_cost_map()[model]
assert model_info["bedrock_output_config_effort_ceiling"] == expected_ceiling
def test_capability_lookups_fall_back_to_base_model_when_regional_entry_lacks_field(
monkeypatch,
):
"""
Regression test: a regional model_cost entry without the capability field
must not shadow a base entry that has it (`get(model) or get(base)` used to
short-circuit on the truthy regional dict and drop the capability).
"""
import litellm
from litellm.llms.bedrock.common_utils import (
bedrock_converse_supports_parallel_tool_use_config,
is_claude_4_5_on_bedrock,
)
base = "anthropic.claude-fallback-test"
regional = f"eu.{base}"
monkeypatch.setitem(litellm.model_cost, regional, {"input_cost_per_token": 1e-06})
monkeypatch.setitem(
litellm.model_cost,
base,
{
"cache_creation_input_token_cost_above_1hr": 1e-05,
"supports_parallel_tool_use_config": True,
},
)
assert is_claude_4_5_on_bedrock(regional) is True
assert bedrock_converse_supports_parallel_tool_use_config(regional) is True

2
tests/test_litellm/test_utils.py
View file

@ -843,6 +843,7 @@ def test_aaamodel_prices_and_context_window_json_is_valid():
"supports_image_input": {"type": "boolean"},
"supports_nova_canvas_image_edit": {"type": "boolean"},
"supports_parallel_function_calling": {"type": "boolean"},
"supports_parallel_tool_use_config": {"type": "boolean"},
"supports_pdf_input": {"type": "boolean"},
"supports_prompt_caching": {"type": "boolean"},
"supports_response_schema": {"type": "boolean"},
@ -869,6 +870,7 @@ def test_aaamodel_prices_and_context_window_json_is_valid():
"type": "string",
"enum": ["low", "medium", "high", "max", "xhigh"],
},
"bedrock_converse_supports_strict_tools": {"type": "boolean"},
"tpm": {"type": "number"},
"provider_specific_entry": {"type": "object"},
"supported_endpoints": {

2
uv.lock generated
View file

@ -3245,7 +3245,7 @@ wheels = [
[[package]]
name = "litellm"
version = "1.90.2"
version = "1.90.3"
source = { editable = "." }
dependencies = [
{ name = "aiohttp" },