vectorvein 0.1.77__tar.gz → 0.1.78__tar.gz

{vectorvein-0.1.77 → vectorvein-0.1.78}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vectorvein
-Version: 0.1.77
+Version: 0.1.78
 Summary: VectorVein python SDK
 Author-Email: Anderson <andersonby@163.com>
 License: MIT

{vectorvein-0.1.77 → vectorvein-0.1.78}/pyproject.toml

@@ -17,7 +17,7 @@ description = "VectorVein python SDK"
 name = "vectorvein"
 readme = "README.md"
 requires-python = ">=3.10"
-version = "0.1.77"
+version = "0.1.78"
 
 [project.license]
 text = "MIT"

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/chat_clients/__init__.py

@@ -21,6 +21,7 @@ from .moonshot_client import MoonshotChatClient, AsyncMoonshotChatClient
 from .deepseek_client import DeepSeekChatClient, AsyncDeepSeekChatClient
 
 from ..types import defaults as defs
+from ..types.llm_parameters import NOT_GIVEN, NotGiven
 from ..types.enums import BackendType, ContextLengthControlType
 from .anthropic_client import AnthropicChatClient, AsyncAnthropicChatClient
 from .utils import format_messages, get_token_counts, get_message_token_counts, ToolCallContentProcessor
@@ -69,7 +70,7 @@ def create_chat_client(
     backend: Literal[BackendType.Anthropic],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -83,7 +84,7 @@ def create_chat_client(
     backend: Literal[BackendType.DeepSeek],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -97,7 +98,7 @@ def create_chat_client(
     backend: Literal[BackendType.Gemini],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -111,7 +112,7 @@ def create_chat_client(
     backend: Literal[BackendType.Groq],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -125,7 +126,7 @@ def create_chat_client(
     backend: Literal[BackendType.Local],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -139,7 +140,7 @@ def create_chat_client(
     backend: Literal[BackendType.MiniMax],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -153,7 +154,7 @@ def create_chat_client(
     backend: Literal[BackendType.Mistral],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -167,7 +168,7 @@ def create_chat_client(
     backend: Literal[BackendType.Moonshot],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -181,7 +182,7 @@ def create_chat_client(
     backend: Literal[BackendType.OpenAI],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -195,7 +196,7 @@ def create_chat_client(
     backend: Literal[BackendType.Qwen],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -209,7 +210,7 @@ def create_chat_client(
     backend: Literal[BackendType.Yi],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -223,7 +224,7 @@ def create_chat_client(
     backend: Literal[BackendType.ZhiPuAI],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -237,7 +238,7 @@ def create_chat_client(
     backend: Literal[BackendType.Baichuan],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -251,7 +252,7 @@ def create_chat_client(
     backend: Literal[BackendType.StepFun],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -265,7 +266,7 @@ def create_chat_client(
     backend: Literal[BackendType.XAI],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -279,7 +280,7 @@ def create_chat_client(
     backend: BackendType,
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -292,7 +293,7 @@ def create_chat_client(
     backend: BackendType,
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -322,7 +323,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Anthropic],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -336,7 +337,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.DeepSeek],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -350,7 +351,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Gemini],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -364,7 +365,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Groq],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -378,7 +379,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Local],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -392,7 +393,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.MiniMax],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -406,7 +407,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Mistral],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -420,7 +421,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Moonshot],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -434,7 +435,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.OpenAI],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -448,7 +449,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Qwen],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -462,7 +463,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Yi],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -476,7 +477,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.ZhiPuAI],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -490,7 +491,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.Baichuan],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -504,7 +505,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.StepFun],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -518,7 +519,7 @@ def create_async_chat_client(
     backend: Literal[BackendType.XAI],
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -532,7 +533,7 @@ def create_async_chat_client(
     backend: BackendType,
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",
@@ -545,7 +546,7 @@ def create_async_chat_client(
     backend: BackendType,
     model: str | None = None,
     stream: bool = False,
-    temperature: float = 0.7,
+    temperature: float | None | NotGiven = NOT_GIVEN,
     context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
     random_endpoint: bool = True,
     endpoint_id: str = "",

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/chat_clients/anthropic_client.py

@@ -157,7 +157,7 @@ class AnthropicChatClient(BaseChatClient):
         self,
         model: str = defs.ANTHROPIC_DEFAULT_MODEL,
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -253,7 +253,7 @@ class AnthropicChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -272,7 +272,7 @@ class AnthropicChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -291,7 +291,7 @@ class AnthropicChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -309,7 +309,7 @@ class AnthropicChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -401,6 +401,8 @@ class AnthropicChatClient(BaseChatClient):
             tool_choice = NOT_GIVEN
         if isinstance(top_p, OpenAINotGiven) or top_p is None:
             top_p = NOT_GIVEN
+        if isinstance(self.temperature, NotGiven) or self.temperature is None:
+            self.temperature = NOT_GIVEN
 
         if messages[0].get("role") == "system":
             system_prompt: str = messages[0]["content"]
@@ -492,7 +494,13 @@ class AnthropicChatClient(BaseChatClient):
                         result["usage"]["total_tokens"] = (
                             result["usage"]["prompt_tokens"] + result["usage"]["completion_tokens"]
                         )
-                        yield ChatCompletionDeltaMessage(usage=Usage(**result["usage"]))
+                        yield ChatCompletionDeltaMessage(
+                            usage=Usage(
+                                prompt_tokens=result["usage"]["prompt_tokens"],
+                                completion_tokens=result["usage"]["completion_tokens"],
+                                total_tokens=result["usage"]["total_tokens"],
+                            )
+                        )
 
             return generator()
         else:
@@ -538,7 +546,7 @@ class AsyncAnthropicChatClient(BaseAsyncChatClient):
         self,
         model: str = defs.ANTHROPIC_DEFAULT_MODEL,
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -633,7 +641,7 @@ class AsyncAnthropicChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -652,7 +660,7 @@ class AsyncAnthropicChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -671,7 +679,7 @@ class AsyncAnthropicChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -689,7 +697,7 @@ class AsyncAnthropicChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -783,6 +791,8 @@ class AsyncAnthropicChatClient(BaseAsyncChatClient):
             tool_choice = NOT_GIVEN
         if isinstance(top_p, OpenAINotGiven) or top_p is None:
             top_p = NOT_GIVEN
+        if isinstance(self.temperature, NotGiven) or self.temperature is None:
+            self.temperature = NOT_GIVEN
 
         if messages[0].get("role") == "system":
             system_prompt = messages[0]["content"]
@@ -874,7 +884,13 @@ class AsyncAnthropicChatClient(BaseAsyncChatClient):
                         result["usage"]["total_tokens"] = (
                             result["usage"]["prompt_tokens"] + result["usage"]["completion_tokens"]
                         )
-                        yield ChatCompletionDeltaMessage(usage=Usage(**result["usage"]))
+                        yield ChatCompletionDeltaMessage(
+                            usage=Usage(
+                                prompt_tokens=result["usage"]["prompt_tokens"],
+                                completion_tokens=result["usage"]["completion_tokens"],
+                                total_tokens=result["usage"]["total_tokens"],
+                            )
+                        )
 
             return generator()
         else:

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/chat_clients/base_client.py

@@ -9,8 +9,8 @@ import httpx
 from openai import OpenAI, AsyncOpenAI, AzureOpenAI, AsyncAzureOpenAI
 from anthropic import (
     Anthropic,
-    AnthropicVertex,
     AsyncAnthropic,
+    AnthropicVertex,
     AsyncAnthropicVertex,
     AnthropicBedrock,
     AsyncAnthropicBedrock,
@@ -39,7 +39,7 @@ class BaseChatClient(ABC):
         self,
         model: str = "",
         stream: bool = False,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -108,7 +108,7 @@ class BaseChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -128,7 +128,7 @@ class BaseChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -148,7 +148,7 @@ class BaseChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -167,7 +167,7 @@ class BaseChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -184,7 +184,7 @@ class BaseChatClient(ABC):
         *,
         messages: list,
         model: str | None = None,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -216,7 +216,7 @@ class BaseAsyncChatClient(ABC):
         self,
         model: str = "",
         stream: bool = False,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -293,7 +293,7 @@ class BaseAsyncChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: list | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -313,7 +313,7 @@ class BaseAsyncChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: list | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -333,7 +333,7 @@ class BaseAsyncChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: list | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -352,7 +352,7 @@ class BaseAsyncChatClient(ABC):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: list | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -369,7 +369,7 @@ class BaseAsyncChatClient(ABC):
         *,
         messages: list,
         model: str | None = None,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: list | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/chat_clients/gemini_client.py

@@ -29,7 +29,7 @@ class GeminiChatClient(BaseChatClient):
         self,
         model: str = defs.GEMINI_DEFAULT_MODEL,
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -63,7 +63,7 @@ class GeminiChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -82,7 +82,7 @@ class GeminiChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -101,7 +101,7 @@ class GeminiChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -119,7 +119,7 @@ class GeminiChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -169,6 +169,10 @@ class GeminiChatClient(BaseChatClient):
         if top_p:
             top_p_params = {"top_p": top_p}
 
+        temperature_params = {}
+        if temperature:
+            temperature_params = {"temperature": temperature}
+
         request_body = {
             "contents": messages,
             "safetySettings": [
@@ -178,8 +182,8 @@ class GeminiChatClient(BaseChatClient):
                 }
             ],
             "generationConfig": {
-                "temperature": self.temperature,
                 "maxOutputTokens": max_tokens,
+                **temperature_params,
                 **top_p_params,
                 **response_format_params,
             },
@@ -277,7 +281,7 @@ class AsyncGeminiChatClient(BaseAsyncChatClient):
         self,
         model: str = defs.GEMINI_DEFAULT_MODEL,
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -311,7 +315,7 @@ class AsyncGeminiChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -330,7 +334,7 @@ class AsyncGeminiChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -349,7 +353,7 @@ class AsyncGeminiChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -367,7 +371,7 @@ class AsyncGeminiChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -417,6 +421,10 @@ class AsyncGeminiChatClient(BaseAsyncChatClient):
         if top_p:
             top_p_params = {"top_p": top_p}
 
+        temperature_params = {}
+        if temperature:
+            temperature_params = {"temperature": temperature}
+
         request_body = {
             "contents": messages,
             "safetySettings": [
@@ -426,8 +434,8 @@ class AsyncGeminiChatClient(BaseAsyncChatClient):
                 }
             ],
             "generationConfig": {
-                "temperature": self.temperature,
                 "maxOutputTokens": max_tokens,
+                **temperature_params,
                 **top_p_params,
                 **response_format_params,
             },

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/chat_clients/minimax_client.py

@@ -51,7 +51,7 @@ class MiniMaxChatClient(BaseChatClient):
         self,
         model: str = defs.MINIMAX_DEFAULT_MODEL,
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -85,7 +85,7 @@ class MiniMaxChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -104,7 +104,7 @@ class MiniMaxChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -123,7 +123,7 @@ class MiniMaxChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -140,7 +140,7 @@ class MiniMaxChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -198,6 +198,10 @@ class MiniMaxChatClient(BaseChatClient):
         else:
             top_p_params = {}
 
+        temperature_params = {}
+        if temperature:
+            temperature_params = {"temperature": temperature}
+
         if max_tokens is None:
             max_output_tokens = self.model_setting.max_output_tokens
             if max_output_tokens is not None:
@@ -223,9 +227,9 @@ class MiniMaxChatClient(BaseChatClient):
             "model": self.model_id,
             "messages": messages,
             "max_tokens": max_tokens,
-            "temperature": self.temperature,
             "stream": self.stream,
             "mask_sensitive_info": False,
+            **temperature_params,
             **top_p_params,
             **tools_params,
             **kwargs,
@@ -298,7 +302,7 @@ class AsyncMiniMaxChatClient(BaseAsyncChatClient):
         self,
         model: str = defs.MINIMAX_DEFAULT_MODEL,
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -332,7 +336,7 @@ class AsyncMiniMaxChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -351,7 +355,7 @@ class AsyncMiniMaxChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -370,7 +374,7 @@ class AsyncMiniMaxChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -388,7 +392,7 @@ class AsyncMiniMaxChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -444,6 +448,10 @@ class AsyncMiniMaxChatClient(BaseAsyncChatClient):
         else:
             top_p_params = {}
 
+        temperature_params = {}
+        if temperature:
+            temperature_params = {"temperature": temperature}
+
         if max_tokens is None:
             max_output_tokens = self.model_setting.max_output_tokens
             if max_output_tokens is not None:
@@ -469,9 +477,9 @@ class AsyncMiniMaxChatClient(BaseAsyncChatClient):
             "model": self.model_id,
             "messages": messages,
             "max_tokens": max_tokens,
-            "temperature": self.temperature,
             "stream": self.stream,
             "mask_sensitive_info": False,
+            **temperature_params,
             **top_p_params,
             **tools_params,
             **kwargs,

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/chat_clients/openai_compatible_client.py

@@ -38,7 +38,7 @@ class OpenAICompatibleChatClient(BaseChatClient):
         self,
         model: str = "",
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -88,7 +88,7 @@ class OpenAICompatibleChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -107,7 +107,7 @@ class OpenAICompatibleChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -126,7 +126,7 @@ class OpenAICompatibleChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -144,7 +144,7 @@ class OpenAICompatibleChatClient(BaseChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -158,6 +158,8 @@ class OpenAICompatibleChatClient(BaseChatClient):
             self.model = model
         if stream is not None:
             self.stream = stream
+        if isinstance(temperature, AnthropicNotGiven):
+            temperature = NOT_GIVEN
         if temperature is not None:
             self.temperature = temperature
         if isinstance(top_p, AnthropicNotGiven):
@@ -233,6 +235,8 @@ class OpenAICompatibleChatClient(BaseChatClient):
                             completion_tokens=chunk.usage.completion_tokens or 0,
                             prompt_tokens=chunk.usage.prompt_tokens or 0,
                             total_tokens=chunk.usage.total_tokens or 0,
+                            prompt_tokens_details=chunk.usage.prompt_tokens_details,
+                            completion_tokens_details=chunk.usage.completion_tokens_details,
                         )
 
                     if len(chunk.choices) == 0:
@@ -279,6 +283,7 @@ class OpenAICompatibleChatClient(BaseChatClient):
 
             result = {
                 "content": response.choices[0].message.content,
+                "reasoning_content": getattr(response.choices[0].message, "reasoning_content", None),
                 "usage": response.usage.model_dump() if response.usage else None,
             }
             if tools:
@@ -305,7 +310,7 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
         self,
         model: str = "",
         stream: bool = True,
-        temperature: float = 0.7,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
         random_endpoint: bool = True,
         endpoint_id: str = "",
@@ -355,7 +360,7 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -374,7 +379,7 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[True],
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -393,7 +398,7 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: bool,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -411,7 +416,7 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
         messages: list,
         model: str | None = None,
         stream: Literal[False] | Literal[True] = False,
-        temperature: float | None = None,
+        temperature: float | None | NotGiven = NOT_GIVEN,
         max_tokens: int | None = None,
         tools: Iterable[ToolParam] | NotGiven = NOT_GIVEN,
         tool_choice: ToolChoice | NotGiven = NOT_GIVEN,
@@ -425,6 +430,8 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
             self.model = model
         if stream is not None:
             self.stream = stream
+        if isinstance(temperature, AnthropicNotGiven):
+            temperature = NOT_GIVEN
         if temperature is not None:
             self.temperature = temperature
         if isinstance(top_p, AnthropicNotGiven):
@@ -500,6 +507,8 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
                             completion_tokens=chunk.usage.completion_tokens or 0,
                             prompt_tokens=chunk.usage.prompt_tokens or 0,
                             total_tokens=chunk.usage.total_tokens or 0,
+                            completion_tokens_details=chunk.usage.completion_tokens_details,
+                            prompt_tokens_details=chunk.usage.prompt_tokens_details,
                         )
 
                     if len(chunk.choices) == 0:
@@ -546,6 +555,7 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
             )
             result = {
                 "content": response.choices[0].message.content,
+                "reasoning_content": getattr(response.choices[0].message, "reasoning_content", None),
                 "usage": response.usage.model_dump() if response.usage else None,
             }
             if tools:

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/types/defaults.py

@@ -39,11 +39,18 @@ MOONSHOT_DEFAULT_MODEL: Final[str] = "moonshot-v1-8k"
 DEEPSEEK_MODELS: Final[Dict[str, Dict[str, Any]]] = {
     "deepseek-chat": {
         "id": "deepseek-chat",
-        "context_length": 128000,
-        "max_output_tokens": 4096,
+        "context_length": 64000,
+        "max_output_tokens": 8192,
         "function_call_available": True,
         "response_format_available": True,
     },
+    "deepseek-reasoner": {
+        "id": "deepseek-reasoner",
+        "context_length": 64000,
+        "max_output_tokens": 8192,
+        "function_call_available": False,
+        "response_format_available": False,
+    },
 }
 DEEPSEEK_DEFAULT_MODEL: Final[str] = "deepseek-chat"
 
@@ -812,4 +819,17 @@ XAI_MODELS: Final[Dict[str, Dict[str, Any]]] = {
         "function_call_available": True,
         "response_format_available": True,
     },
+    "grok-2-latest": {
+        "id": "grok-2-latest",
+        "context_length": 131072,
+        "function_call_available": True,
+        "response_format_available": True,
+    },
+    "grok-2-vision-latest": {
+        "id": "grok-2-vision-latest",
+        "context_length": 32768,
+        "function_call_available": True,
+        "response_format_available": True,
+        "native_multimodal": True,
+    },
 }

{vectorvein-0.1.77 → vectorvein-0.1.78}/src/vectorvein/types/llm_parameters.py

@@ -13,6 +13,7 @@ from openai._types import NotGiven as OpenAINotGiven
 from openai._types import NOT_GIVEN as OPENAI_NOT_GIVEN
 from openai.types.chat.chat_completion_chunk import ChoiceDeltaToolCall
 from openai.types.chat.chat_completion_tool_param import ChatCompletionToolParam
+from openai.types.completion_usage import CompletionTokensDetails, PromptTokensDetails
 from openai.types.chat.chat_completion_message_tool_call import ChatCompletionMessageToolCall
 from openai.types.chat.chat_completion_stream_options_param import ChatCompletionStreamOptionsParam
 from openai.types.chat.chat_completion_tool_choice_option_param import ChatCompletionToolChoiceOptionParam
@@ -80,13 +81,23 @@ class BackendSettings(BaseModel):
 
 class Usage(BaseModel):
     completion_tokens: int
+
     prompt_tokens: int
+
     total_tokens: int
 
+    completion_tokens_details: Optional[CompletionTokensDetails] = None
+    """Breakdown of tokens used in a completion."""
+
+    prompt_tokens_details: Optional[PromptTokensDetails] = None
+    """Breakdown of tokens used in the prompt."""
+
 
 class ChatCompletionMessage(BaseModel):
     content: Optional[str] = None
 
+    reasoning_content: Optional[str] = None
+
     tool_calls: Optional[List[ChatCompletionMessageToolCall]] = None
     """The tool calls generated by the model, such as function calls."""
 
@@ -98,6 +109,8 @@ class ChatCompletionMessage(BaseModel):
 class ChatCompletionDeltaMessage(BaseModel):
     content: Optional[str] = None
 
+    reasoning_content: Optional[str] = None
+
     tool_calls: Optional[List[ChoiceDeltaToolCall]] = None
     """The tool calls generated by the model, such as function calls."""