diff --git a/python/packages/autogen-agentchat/src/autogen_agentchat/agents/_assistant_agent.py b/python/packages/autogen-agentchat/src/autogen_agentchat/agents/_assistant_agent.py
index 65e01817f59b..5b0575d7c474 100644
--- a/python/packages/autogen-agentchat/src/autogen_agentchat/agents/_assistant_agent.py
+++ b/python/packages/autogen-agentchat/src/autogen_agentchat/agents/_assistant_agent.py
@@ -13,7 +13,7 @@
     Sequence,
 )
 
-from autogen_core import CancellationToken, Component, ComponentModel, FunctionCall
+from autogen_core import CancellationToken, Component, ComponentModel, FunctionCall, FunctionCalls
 from autogen_core.memory import Memory
 from autogen_core.model_context import (
     ChatCompletionContext,
@@ -401,9 +401,10 @@ async def on_messages_stream(
             return
 
         # Process tool calls.
-        assert isinstance(model_result.content, list) and all(
-            isinstance(item, FunctionCall) for item in model_result.content
+        assert isinstance(model_result.content, FunctionCalls) and all(
+            isinstance(item, FunctionCall) or isinstance(item, str) for item in model_result.content.function_calls
         )
+
         tool_call_msg = ToolCallRequestEvent(
             content=model_result.content, source=self.name, models_usage=model_result.usage
         )
@@ -412,9 +413,15 @@ async def on_messages_stream(
         inner_messages.append(tool_call_msg)
         yield tool_call_msg
 
+        function_calls = model_result.content.function_calls
+
         # Execute the tool calls.
         exec_results = await asyncio.gather(
-            *[self._execute_tool_call(call, cancellation_token) for call in model_result.content]
+            *[
+                self._execute_tool_call(call, cancellation_token)
+                for call in function_calls
+                if isinstance(call, FunctionCall)
+            ]
         )
         tool_call_result_msg = ToolCallExecutionEvent(content=exec_results, source=self.name)
         event_logger.debug(tool_call_result_msg)
@@ -423,7 +430,7 @@ async def on_messages_stream(
         yield tool_call_result_msg
 
         # Correlate tool call results with tool calls.
-        tool_calls = [call for call in model_result.content if call.name not in self._handoffs]
+        tool_calls = [call for call in function_calls if call.name not in self._handoffs]
         tool_call_results: List[FunctionExecutionResult] = []
         for tool_call in tool_calls:
             found = False
@@ -436,7 +443,7 @@ async def on_messages_stream(
                 raise RuntimeError(f"Tool call result not found for call id: {tool_call.id}")
 
         # Detect handoff requests.
-        handoff_reqs = [call for call in model_result.content if call.name in self._handoffs]
+        handoff_reqs = [call for call in function_calls if call.name in self._handoffs]
         if len(handoff_reqs) > 0:
             handoffs = [self._handoffs[call.name] for call in handoff_reqs]
             if len(handoffs) > 1:
@@ -451,7 +458,12 @@ async def on_messages_stream(
             # Current context for handoff.
             handoff_context: List[LLMMessage] = []
             if len(tool_calls) > 0:
-                handoff_context.append(AssistantMessage(content=tool_calls, source=self.name))
+                handoff_context.append(
+                    AssistantMessage(
+                        content=FunctionCalls(function_calls=tool_calls, thought=model_result.content.thought),
+                        source=self.name,
+                    )
+                )
                 handoff_context.append(FunctionExecutionResultMessage(content=tool_call_results))
             # Return the output messages to signal the handoff.
             yield Response(
diff --git a/python/packages/autogen-agentchat/src/autogen_agentchat/messages.py b/python/packages/autogen-agentchat/src/autogen_agentchat/messages.py
index 25d9e732d335..7f1891076bce 100644
--- a/python/packages/autogen-agentchat/src/autogen_agentchat/messages.py
+++ b/python/packages/autogen-agentchat/src/autogen_agentchat/messages.py
@@ -7,7 +7,7 @@ class and includes specific fields relevant to the type of message being sent.
 from abc import ABC
 from typing import List, Literal
 
-from autogen_core import FunctionCall, Image
+from autogen_core import FunctionCalls, Image
 from autogen_core.memory import MemoryContent
 from autogen_core.models import FunctionExecutionResult, LLMMessage, RequestUsage
 from pydantic import BaseModel, ConfigDict, Field
@@ -83,7 +83,7 @@ class HandoffMessage(BaseChatMessage):
 class ToolCallRequestEvent(BaseAgentEvent):
     """An event signaling a request to use tools."""
 
-    content: List[FunctionCall]
+    content: FunctionCalls
     """The tool calls."""
 
     type: Literal["ToolCallRequestEvent"] = "ToolCallRequestEvent"
diff --git a/python/packages/autogen-agentchat/tests/test_assistant_agent.py b/python/packages/autogen-agentchat/tests/test_assistant_agent.py
index 7b281e39f6f5..fade29b3ee99 100644
--- a/python/packages/autogen-agentchat/tests/test_assistant_agent.py
+++ b/python/packages/autogen-agentchat/tests/test_assistant_agent.py
@@ -2,6 +2,7 @@
 import json
 import logging
 from typing import Any, AsyncGenerator, List
+from unittest.mock import AsyncMock
 
 import pytest
 from autogen_agentchat import EVENT_LOGGER_NAME
@@ -17,10 +18,10 @@
     ToolCallRequestEvent,
     ToolCallSummaryMessage,
 )
-from autogen_core import FunctionCall, Image
+from autogen_core import FunctionCall, FunctionCalls, Image
 from autogen_core.memory import ListMemory, Memory, MemoryContent, MemoryMimeType, MemoryQueryResult
 from autogen_core.model_context import BufferedChatCompletionContext
-from autogen_core.models import FunctionExecutionResult, LLMMessage
+from autogen_core.models import CreateResult, FunctionExecutionResult, LLMMessage, RequestUsage
 from autogen_core.models._model_client import ModelFamily
 from autogen_core.tools import FunctionTool
 from autogen_ext.models.openai import OpenAIChatCompletionClient
@@ -176,6 +177,51 @@ async def test_run_with_tools(monkeypatch: pytest.MonkeyPatch) -> None:
     assert state == state2
 
 
+@pytest.mark.asyncio
+async def test_run_stream_with_thought() -> None:
+    mocked_model = AsyncMock()
+    mocked_model.create.side_effect = [
+        CreateResult(
+            finish_reason="function_calls",
+            content=FunctionCalls(
+                function_calls=[FunctionCall(id="call_foo", name="echo", arguments=json.dumps({"input": "foo"}))],
+                thought="going to say foo!",
+            ),
+            usage=RequestUsage(prompt_tokens=10, completion_tokens=5),
+            cached=False,
+        ),
+        CreateResult(
+            finish_reason="stop",
+            content="ok!",
+            usage=RequestUsage(prompt_tokens=10, completion_tokens=5),
+            cached=False,
+        ),
+        CreateResult(
+            finish_reason="stop",
+            content="TERMINATE",
+            usage=RequestUsage(prompt_tokens=10, completion_tokens=5),
+            cached=False,
+        ),
+    ]
+
+    agent = AssistantAgent(
+        "thoughtfull_tool_use_agent",
+        model_client=mocked_model,
+        tools=[
+            FunctionTool(_echo_function, description="Echo", name="echo"),
+        ],
+    )
+
+    streamed_messages = [item async for item in agent.run_stream(task="prompt")]
+    assert len(streamed_messages) == 5
+    assert isinstance(streamed_messages[0], TextMessage)
+    assert isinstance(streamed_messages[1], ToolCallRequestEvent)
+    assert streamed_messages[1].content.thought == "going to say foo!"
+    assert isinstance(streamed_messages[2], ToolCallExecutionEvent)
+    assert isinstance(streamed_messages[3], ToolCallSummaryMessage)
+    assert isinstance(streamed_messages[4], TaskResult)
+
+
 @pytest.mark.asyncio
 async def test_run_with_tools_and_reflection(monkeypatch: pytest.MonkeyPatch) -> None:
     model = "gpt-4o-2024-05-13"
@@ -374,7 +420,7 @@ async def test_run_with_parallel_tools(monkeypatch: pytest.MonkeyPatch) -> None:
     assert isinstance(result.messages[0], TextMessage)
     assert result.messages[0].models_usage is None
     assert isinstance(result.messages[1], ToolCallRequestEvent)
-    assert result.messages[1].content == [
+    assert result.messages[1].content.function_calls == [
         FunctionCall(id="1", arguments=r'{"input": "task1"}', name="_pass_function"),
         FunctionCall(id="2", arguments=r'{"input": "task2"}', name="_pass_function"),
         FunctionCall(id="3", arguments=r'{"input": "task3"}', name="_echo_function"),
diff --git a/python/packages/autogen-agentchat/tests/test_group_chat.py b/python/packages/autogen-agentchat/tests/test_group_chat.py
index c04d7344029f..b8d433f2f3b1 100644
--- a/python/packages/autogen-agentchat/tests/test_group_chat.py
+++ b/python/packages/autogen-agentchat/tests/test_group_chat.py
@@ -29,7 +29,7 @@
 from autogen_agentchat.teams._group_chat._selector_group_chat import SelectorGroupChatManager
 from autogen_agentchat.teams._group_chat._swarm_group_chat import SwarmGroupChatManager
 from autogen_agentchat.ui import Console
-from autogen_core import AgentId, CancellationToken, FunctionCall
+from autogen_core import AgentId, CancellationToken, FunctionCall, FunctionCalls
 from autogen_core.models import (
     AssistantMessage,
     FunctionExecutionResult,
@@ -220,6 +220,7 @@ async def test_round_robin_group_chat(monkeypatch: pytest.MonkeyPatch) -> None:
         result_2 = await team.run(
             task=MultiModalMessage(content=["Write a program that prints 'Hello, world!'"], source="user")
         )
+        assert isinstance(result_2.messages[0].content, list)
         assert result.messages[0].content == result_2.messages[0].content[0]
         assert result.messages[1:] == result_2.messages[1:]
 
@@ -1065,10 +1066,12 @@ async def test_swarm_with_parallel_tool_calls(monkeypatch: pytest.MonkeyPatch) -
     expected_handoff_context: List[LLMMessage] = [
         AssistantMessage(
             source="agent1",
-            content=[
-                FunctionCall(id="1", name="tool1", arguments="{}"),
-                FunctionCall(id="2", name="tool2", arguments="{}"),
-            ],
+            content=FunctionCalls(
+                function_calls=[
+                    FunctionCall(id="1", name="tool1", arguments="{}"),
+                    FunctionCall(id="2", name="tool2", arguments="{}"),
+                ]
+            ),
         ),
         FunctionExecutionResultMessage(
             content=[
diff --git a/python/packages/autogen-core/docs/src/user-guide/agentchat-user-guide/migration-guide.md b/python/packages/autogen-core/docs/src/user-guide/agentchat-user-guide/migration-guide.md
index a0beff3e2f8c..e796d19c313c 100644
--- a/python/packages/autogen-core/docs/src/user-guide/agentchat-user-guide/migration-guide.md
+++ b/python/packages/autogen-core/docs/src/user-guide/agentchat-user-guide/migration-guide.md
@@ -627,7 +627,7 @@ from autogen_agentchat.messages import (
     ToolCallRequestEvent,
     ToolCallSummaryMessage,
 )
-from autogen_core import FunctionCall, Image
+from autogen_core import FunctionCall, FunctionCalls, Image
 from autogen_core.models import FunctionExecutionResult
 
 
@@ -660,7 +660,7 @@ def convert_to_v02_message(
                 raise ValueError(f"Invalid multimodal message content: {modal}")
     elif isinstance(message, ToolCallRequestEvent):
         v02_message = {"tool_calls": [], "role": "assistant", "content": None, "name": message.source}
-        for tool_call in message.content:
+        for tool_call in message.content.function_calls:
             v02_message["tool_calls"].append(
                 {
                     "id": tool_call.id,
@@ -697,7 +697,7 @@ def convert_to_v04_message(message: Dict[str, Any]) -> AgentEvent | ChatMessage:
                     arguments=tool_call["function"]["args"],
                 )
             )
-        return ToolCallRequestEvent(source=message["name"], content=tool_calls)
+        return ToolCallRequestEvent(source=message["name"], content=FunctionCalls(function_calls=tool_calls))
     elif "tool_responses" in message:
         tool_results: List[FunctionExecutionResult] = []
         for tool_response in message["tool_responses"]:
diff --git a/python/packages/autogen-core/docs/src/user-guide/core-user-guide/design-patterns/handoffs.ipynb b/python/packages/autogen-core/docs/src/user-guide/core-user-guide/design-patterns/handoffs.ipynb
index 31885a10e977..7e6c46f69297 100644
--- a/python/packages/autogen-core/docs/src/user-guide/core-user-guide/design-patterns/handoffs.ipynb
+++ b/python/packages/autogen-core/docs/src/user-guide/core-user-guide/design-patterns/handoffs.ipynb
@@ -58,6 +58,7 @@
     "\n",
     "from autogen_core import (\n",
     "    FunctionCall,\n",
+    "    FunctionCalls,\n",
     "    MessageContext,\n",
     "    RoutedAgent,\n",
     "    SingleThreadedAgentRuntime,\n",
@@ -173,11 +174,13 @@
     "        )\n",
     "        print(f\"{'-'*80}\\n{self.id.type}:\\n{llm_result.content}\", flush=True)\n",
     "        # Process the LLM result.\n",
-    "        while isinstance(llm_result.content, list) and all(isinstance(m, FunctionCall) for m in llm_result.content):\n",
+    "        while isinstance(llm_result.content, FunctionCalls) and all(\n",
+    "            isinstance(m, FunctionCall) for m in llm_result.content.function_calls\n",
+    "        ):\n",
     "            tool_call_results: List[FunctionExecutionResult] = []\n",
     "            delegate_targets: List[Tuple[str, UserTask]] = []\n",
     "            # Process each function call.\n",
-    "            for call in llm_result.content:\n",
+    "            for call in llm_result.content.function_calls:\n",
     "                arguments = json.loads(call.arguments)\n",
     "                if call.name in self._tools:\n",
     "                    # Execute the tool directly.\n",
@@ -190,7 +193,7 @@
     "                    topic_type = self._delegate_tools[call.name].return_value_as_string(result)\n",
     "                    # Create the context for the delegate agent, including the function call and the result.\n",
     "                    delegate_messages = list(message.context) + [\n",
-    "                        AssistantMessage(content=[call], source=self.id.type),\n",
+    "                        AssistantMessage(content=FunctionCalls(function_calls=[call]), source=self.id.type),\n",
     "                        FunctionExecutionResultMessage(\n",
     "                            content=[\n",
     "                                FunctionExecutionResult(\n",
diff --git a/python/packages/autogen-core/src/autogen_core/__init__.py b/python/packages/autogen-core/src/autogen_core/__init__.py
index 0198544ca61e..1d3af45e19da 100644
--- a/python/packages/autogen-core/src/autogen_core/__init__.py
+++ b/python/packages/autogen-core/src/autogen_core/__init__.py
@@ -62,7 +62,7 @@
 from ._topic import TopicId
 from ._type_prefix_subscription import TypePrefixSubscription
 from ._type_subscription import TypeSubscription
-from ._types import FunctionCall
+from ._types import FunctionCall, FunctionCalls
 
 EVENT_LOGGER_NAME = EVENT_LOGGER_NAME_ALIAS
 """The name of the logger used for structured events."""
@@ -107,6 +107,7 @@
     "event",
     "rpc",
     "FunctionCall",
+    "FunctionCalls",
     "TypeSubscription",
     "DefaultSubscription",
     "DefaultTopicId",
diff --git a/python/packages/autogen-core/src/autogen_core/_types.py b/python/packages/autogen-core/src/autogen_core/_types.py
index 5e3850ffae8b..ff54a76ce86f 100644
--- a/python/packages/autogen-core/src/autogen_core/_types.py
+++ b/python/packages/autogen-core/src/autogen_core/_types.py
@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 from dataclasses import dataclass
+from typing import List, Optional
 
 
 @dataclass
@@ -10,3 +11,9 @@ class FunctionCall:
     arguments: str
     # Function to call
     name: str
+
+
+@dataclass
+class FunctionCalls:
+    function_calls: List[FunctionCall]
+    thought: Optional[str] = None
diff --git a/python/packages/autogen-core/src/autogen_core/model_context/_head_and_tail_chat_completion_context.py b/python/packages/autogen-core/src/autogen_core/model_context/_head_and_tail_chat_completion_context.py
index a37d5927b19f..982084a43370 100644
--- a/python/packages/autogen-core/src/autogen_core/model_context/_head_and_tail_chat_completion_context.py
+++ b/python/packages/autogen-core/src/autogen_core/model_context/_head_and_tail_chat_completion_context.py
@@ -4,7 +4,7 @@
 from typing_extensions import Self
 
 from .._component_config import Component
-from .._types import FunctionCall
+from .._types import FunctionCall, FunctionCalls
 from ..models import AssistantMessage, FunctionExecutionResultMessage, LLMMessage, UserMessage
 from ._chat_completion_context import ChatCompletionContext
 
@@ -45,8 +45,8 @@ async def get_messages(self) -> List[LLMMessage]:
         if (
             head_messages
             and isinstance(head_messages[-1], AssistantMessage)
-            and isinstance(head_messages[-1].content, list)
-            and all(isinstance(item, FunctionCall) for item in head_messages[-1].content)
+            and isinstance(head_messages[-1].content, FunctionCalls)
+            and all(isinstance(item, FunctionCall) for item in head_messages[-1].content.function_calls)
         ):
             # Remove the last message from the head.
             head_messages = head_messages[:-1]
diff --git a/python/packages/autogen-core/src/autogen_core/models/_types.py b/python/packages/autogen-core/src/autogen_core/models/_types.py
index a3d6af1edde4..5944730645f0 100644
--- a/python/packages/autogen-core/src/autogen_core/models/_types.py
+++ b/python/packages/autogen-core/src/autogen_core/models/_types.py
@@ -4,7 +4,7 @@
 from pydantic import BaseModel, Field
 from typing_extensions import Annotated
 
-from .. import FunctionCall, Image
+from .. import FunctionCalls, Image
 
 
 class SystemMessage(BaseModel):
@@ -22,7 +22,7 @@ class UserMessage(BaseModel):
 
 
 class AssistantMessage(BaseModel):
-    content: Union[str, List[FunctionCall]]
+    content: Union[str, FunctionCalls]
 
     # Name of the agent that sent this message
     source: str
@@ -70,7 +70,7 @@ class ChatCompletionTokenLogprob(BaseModel):
 
 class CreateResult(BaseModel):
     finish_reason: FinishReasons
-    content: Union[str, List[FunctionCall]]
+    content: Union[str, FunctionCalls]
     usage: RequestUsage
     cached: bool
     logprobs: Optional[List[ChatCompletionTokenLogprob] | None] = None
diff --git a/python/packages/autogen-core/src/autogen_core/tool_agent/_caller_loop.py b/python/packages/autogen-core/src/autogen_core/tool_agent/_caller_loop.py
index 2353ca184f86..a12634907434 100644
--- a/python/packages/autogen-core/src/autogen_core/tool_agent/_caller_loop.py
+++ b/python/packages/autogen-core/src/autogen_core/tool_agent/_caller_loop.py
@@ -1,7 +1,7 @@
 import asyncio
 from typing import List
 
-from .. import AgentId, AgentRuntime, BaseAgent, CancellationToken, FunctionCall
+from .. import AgentId, AgentRuntime, BaseAgent, CancellationToken, FunctionCall, FunctionCalls
 from ..models import (
     AssistantMessage,
     ChatCompletionClient,
@@ -43,7 +43,9 @@ async def tool_agent_caller_loop(
     generated_messages.append(AssistantMessage(content=response.content, source=caller_source))
 
     # Keep iterating until the model stops generating tool calls.
-    while isinstance(response.content, list) and all(isinstance(item, FunctionCall) for item in response.content):
+    while isinstance(response.content, FunctionCalls) and all(
+        isinstance(item, FunctionCall) for item in response.content.function_calls
+    ):
         # Execute functions called by the model by sending messages to tool agent.
         results: List[FunctionExecutionResult | BaseException] = await asyncio.gather(
             *[
@@ -52,7 +54,7 @@ async def tool_agent_caller_loop(
                     recipient=tool_agent_id,
                     cancellation_token=cancellation_token,
                 )
-                for call in response.content
+                for call in response.content.function_calls
             ],
             return_exceptions=True,
         )
diff --git a/python/packages/autogen-core/tests/test_memory.py b/python/packages/autogen-core/tests/test_memory.py
index 04054e1b2250..bab41fb2aed6 100644
--- a/python/packages/autogen-core/tests/test_memory.py
+++ b/python/packages/autogen-core/tests/test_memory.py
@@ -115,6 +115,7 @@ async def test_list_memory_update_context() -> None:
     context_messages = await context.get_messages()
     assert len(results.memories.results) == 2
     assert len(context_messages) == 1
+    assert isinstance(context_messages[0].content, str)
     assert "test1" in context_messages[0].content
     assert "test2" in context_messages[0].content
 
diff --git a/python/packages/autogen-core/tests/test_tool_agent.py b/python/packages/autogen-core/tests/test_tool_agent.py
index 85fcd3892c97..7023be6654a2 100644
--- a/python/packages/autogen-core/tests/test_tool_agent.py
+++ b/python/packages/autogen-core/tests/test_tool_agent.py
@@ -3,7 +3,7 @@
 from typing import Any, AsyncGenerator, List, Mapping, Optional, Sequence, Union
 
 import pytest
-from autogen_core import AgentId, CancellationToken, FunctionCall, SingleThreadedAgentRuntime
+from autogen_core import AgentId, CancellationToken, FunctionCall, FunctionCalls, SingleThreadedAgentRuntime
 from autogen_core.models import (
     AssistantMessage,
     ChatCompletionClient,
@@ -101,7 +101,9 @@ async def create(
         ) -> CreateResult:
             if len(messages) == 1:
                 return CreateResult(
-                    content=[FunctionCall(id="1", name="pass", arguments=json.dumps({"input": "test"}))],
+                    content=FunctionCalls(
+                        function_calls=[FunctionCall(id="1", name="pass", arguments=json.dumps({"input": "test"}))]
+                    ),
                     finish_reason="stop",
                     usage=RequestUsage(prompt_tokens=0, completion_tokens=0),
                     cached=False,
diff --git a/python/packages/autogen-ext/src/autogen_ext/agents/file_surfer/_file_surfer.py b/python/packages/autogen-ext/src/autogen_ext/agents/file_surfer/_file_surfer.py
index 0f389313057a..662766eddbb6 100644
--- a/python/packages/autogen-ext/src/autogen_ext/agents/file_surfer/_file_surfer.py
+++ b/python/packages/autogen-ext/src/autogen_ext/agents/file_surfer/_file_surfer.py
@@ -9,7 +9,7 @@
     MultiModalMessage,
     TextMessage,
 )
-from autogen_core import CancellationToken, FunctionCall
+from autogen_core import CancellationToken, FunctionCall, FunctionCalls
 from autogen_core.models import (
     AssistantMessage,
     ChatCompletionClient,
@@ -143,8 +143,10 @@ async def _generate_reply(self, cancellation_token: CancellationToken) -> Tuple[
             # Answer directly.
             return False, response
 
-        elif isinstance(response, list) and all(isinstance(item, FunctionCall) for item in response):
-            function_calls = response
+        elif isinstance(response, FunctionCalls) and all(
+            isinstance(item, FunctionCall) for item in response.function_calls
+        ):
+            function_calls = response.function_calls
             for function_call in function_calls:
                 tool_name = function_call.name
 
diff --git a/python/packages/autogen-ext/src/autogen_ext/agents/openai/_openai_assistant_agent.py b/python/packages/autogen-ext/src/autogen_ext/agents/openai/_openai_assistant_agent.py
index d738a5fae894..30a04e7cbd07 100644
--- a/python/packages/autogen-ext/src/autogen_ext/agents/openai/_openai_assistant_agent.py
+++ b/python/packages/autogen-ext/src/autogen_ext/agents/openai/_openai_assistant_agent.py
@@ -32,7 +32,7 @@
     ToolCallExecutionEvent,
     ToolCallRequestEvent,
 )
-from autogen_core import CancellationToken, FunctionCall
+from autogen_core import CancellationToken, FunctionCall, FunctionCalls
 from autogen_core.models._model_client import ChatCompletionClient
 from autogen_core.models._types import FunctionExecutionResult
 from autogen_core.tools import FunctionTool, Tool
@@ -395,7 +395,7 @@ async def on_messages_stream(
                         )
 
                 # Add tool call message to inner messages
-                tool_call_msg = ToolCallRequestEvent(source=self.name, content=tool_calls)
+                tool_call_msg = ToolCallRequestEvent(source=self.name, content=FunctionCalls(function_calls=tool_calls))
                 inner_messages.append(tool_call_msg)
                 event_logger.debug(tool_call_msg)
                 yield tool_call_msg
diff --git a/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_multimodal_web_surfer.py b/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_multimodal_web_surfer.py
index f90dc01cdda2..db4467fe3510 100644
--- a/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_multimodal_web_surfer.py
+++ b/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_multimodal_web_surfer.py
@@ -24,7 +24,7 @@
 from autogen_agentchat.agents import BaseChatAgent
 from autogen_agentchat.base import Response
 from autogen_agentchat.messages import AgentEvent, ChatMessage, MultiModalMessage, TextMessage
-from autogen_core import EVENT_LOGGER_NAME, CancellationToken, Component, ComponentModel, FunctionCall
+from autogen_core import EVENT_LOGGER_NAME, CancellationToken, Component, ComponentModel, FunctionCall, FunctionCalls
 from autogen_core import Image as AGImage
 from autogen_core.models import (
     AssistantMessage,
@@ -541,9 +541,11 @@ async def _generate_reply(self, cancellation_token: CancellationToken) -> UserCo
             # Answer directly
             self.inner_messages.append(TextMessage(content=message, source=self.name))
             return message
-        elif isinstance(message, list):
+        elif isinstance(message, FunctionCalls):
             # Take an action
-            return await self._execute_tool(message, rects, tool_names, cancellation_token=cancellation_token)
+            return await self._execute_tool(
+                message.function_calls, rects, tool_names, cancellation_token=cancellation_token
+            )
         else:
             # Not sure what happened here
             raise AssertionError(f"Unknown response format '{message}'")
diff --git a/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_types.py b/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_types.py
index d626b086961d..29b77acc86e5 100644
--- a/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_types.py
+++ b/python/packages/autogen-ext/src/autogen_ext/agents/web_surfer/_types.py
@@ -1,10 +1,10 @@
 from typing import Any, Dict, List, TypedDict, Union
 
-from autogen_core import FunctionCall, Image
+from autogen_core import FunctionCalls, Image
 from autogen_core.models import FunctionExecutionResult
 
 UserContent = Union[str, List[Union[str, Image]]]
-AssistantContent = Union[str, List[FunctionCall]]
+AssistantContent = Union[str, FunctionCalls]
 FunctionExecutionContent = List[FunctionExecutionResult]
 SystemContent = str
 
diff --git a/python/packages/autogen-ext/src/autogen_ext/models/openai/_openai_client.py b/python/packages/autogen-ext/src/autogen_ext/models/openai/_openai_client.py
index 79c13442c7dc..09b3390c491e 100644
--- a/python/packages/autogen-ext/src/autogen_ext/models/openai/_openai_client.py
+++ b/python/packages/autogen-ext/src/autogen_ext/models/openai/_openai_client.py
@@ -27,6 +27,7 @@
     CancellationToken,
     Component,
     FunctionCall,
+    FunctionCalls,
     Image,
     MessageHandlerContext,
 )
@@ -196,9 +197,9 @@ def assistant_message_to_oai(
     message: AssistantMessage,
 ) -> ChatCompletionAssistantMessageParam:
     assert_valid_name(message.source)
-    if isinstance(message.content, list):
+    if isinstance(message.content, FunctionCalls):
         return ChatCompletionAssistantMessageParam(
-            tool_calls=[func_call_to_oai(x) for x in message.content],
+            tool_calls=[func_call_to_oai(x) for x in message.content.function_calls],
             role="assistant",
             name=message.source,
         )
@@ -549,7 +550,7 @@ async def create(
 
         # Detect whether it is a function call or not.
         # We don't rely on choice.finish_reason as it is not always accurate, depending on the API used.
-        content: Union[str, List[FunctionCall]]
+        content: Union[str, FunctionCalls]
         if choice.message.function_call is not None:
             raise ValueError("function_call is deprecated and is not supported by this model client.")
         elif choice.message.tool_calls is not None:
@@ -567,14 +568,16 @@ async def create(
                     stacklevel=2,
                 )
             # NOTE: If OAI response type changes, this will need to be updated
-            content = [
-                FunctionCall(
-                    id=x.id,
-                    arguments=x.function.arguments,
-                    name=normalize_name(x.function.name),
-                )
-                for x in choice.message.tool_calls
-            ]
+            content = FunctionCalls(
+                function_calls=[
+                    FunctionCall(
+                        id=x.id,
+                        arguments=x.function.arguments,
+                        name=normalize_name(x.function.name),
+                    )
+                    for x in choice.message.tool_calls
+                ]
+            )
             finish_reason = "tool_calls"
         else:
             finish_reason = choice.finish_reason
@@ -783,7 +786,7 @@ async def create_stream(
         if stop_reason == "function_call":
             raise ValueError("Function calls are not supported in this context")
 
-        content: Union[str, List[FunctionCall]]
+        content: Union[str, FunctionCalls]
         if len(content_deltas) > 1:
             content = "".join(content_deltas)
             if chunk and chunk.usage:
@@ -797,7 +800,7 @@ async def create_stream(
             #     # value = json.dumps(tool_call)
             #     # completion_tokens += count_token(value, model=model)
             #     completion_tokens += 0
-            content = list(full_tool_calls.values())
+            content = FunctionCalls(function_calls=list(full_tool_calls.values()))
 
         usage = RequestUsage(
             prompt_tokens=prompt_tokens,
diff --git a/python/packages/autogen-ext/src/autogen_ext/models/semantic_kernel/_sk_chat_completion_adapter.py b/python/packages/autogen-ext/src/autogen_ext/models/semantic_kernel/_sk_chat_completion_adapter.py
index 9f501fbbffcf..e6b910fe425f 100644
--- a/python/packages/autogen-ext/src/autogen_ext/models/semantic_kernel/_sk_chat_completion_adapter.py
+++ b/python/packages/autogen-ext/src/autogen_ext/models/semantic_kernel/_sk_chat_completion_adapter.py
@@ -1,7 +1,7 @@
 import json
 from typing import Any, Literal, Mapping, Optional, Sequence
 
-from autogen_core import FunctionCall
+from autogen_core import FunctionCall, FunctionCalls
 from autogen_core._cancellation_token import CancellationToken
 from autogen_core.models import (
     ChatCompletionClient,
@@ -361,9 +361,9 @@ async def create(
         self._total_completion_tokens += completion_tokens
 
         # Process content based on whether there are tool calls
-        content: Union[str, list[FunctionCall]]
+        content: Union[str, FunctionCalls]
         if any(isinstance(item, FunctionCallContent) for item in result[0].items):
-            content = self._process_tool_calls(result[0])
+            content = FunctionCalls(function_calls=self._process_tool_calls(result[0]))
             finish_reason: Literal["function_calls", "stop"] = "function_calls"
         else:
             content = result[0].content
@@ -436,7 +436,7 @@ async def create_stream(
                 if any(isinstance(item, FunctionCallContent) for item in msg.items):
                     function_calls = self._process_tool_calls(msg)
                     yield CreateResult(
-                        content=function_calls,
+                        content=FunctionCalls(function_calls=function_calls),
                         finish_reason="function_calls",
                         usage=RequestUsage(prompt_tokens=prompt_tokens, completion_tokens=completion_tokens),
                         cached=False,
diff --git a/python/packages/autogen-ext/tests/models/test_openai_model_client.py b/python/packages/autogen-ext/tests/models/test_openai_model_client.py
index d629cdc428a5..7957c7f2bfde 100644
--- a/python/packages/autogen-ext/tests/models/test_openai_model_client.py
+++ b/python/packages/autogen-ext/tests/models/test_openai_model_client.py
@@ -5,7 +5,7 @@
 from unittest.mock import MagicMock
 
 import pytest
-from autogen_core import CancellationToken, FunctionCall, Image
+from autogen_core import CancellationToken, FunctionCall, FunctionCalls, Image
 from autogen_core.models import (
     AssistantMessage,
     CreateResult,
@@ -610,7 +610,10 @@ async def test_tool_calling(monkeypatch: pytest.MonkeyPatch) -> None:
 
     # Single tool call
     create_result = await model_client.create(messages=[UserMessage(content="Hello", source="user")], tools=[pass_tool])
-    assert create_result.content == [FunctionCall(id="1", arguments=r'{"input": "task"}', name="_pass_function")]
+    assert isinstance(create_result.content, FunctionCalls)
+    assert create_result.content.function_calls == [
+        FunctionCall(id="1", arguments=r'{"input": "task"}', name="_pass_function")
+    ]
     # Verify that the tool schema was passed to the model client.
     kwargs = mock.calls[0]
     assert kwargs["tools"] == [{"function": pass_tool.schema, "type": "function"}]
@@ -621,7 +624,8 @@ async def test_tool_calling(monkeypatch: pytest.MonkeyPatch) -> None:
     create_result = await model_client.create(
         messages=[UserMessage(content="Hello", source="user")], tools=[pass_tool, fail_tool, echo_tool]
     )
-    assert create_result.content == [
+    assert isinstance(create_result.content, FunctionCalls)
+    assert create_result.content.function_calls == [
         FunctionCall(id="1", arguments=r'{"input": "task"}', name="_pass_function"),
         FunctionCall(id="2", arguments=r'{"input": "task"}', name="_fail_function"),
         FunctionCall(id="3", arguments=r'{"input": "task"}', name="_echo_function"),
@@ -641,7 +645,10 @@ async def test_tool_calling(monkeypatch: pytest.MonkeyPatch) -> None:
         create_result = await model_client.create(
             messages=[UserMessage(content="Hello", source="user")], tools=[pass_tool]
         )
-        assert create_result.content == [FunctionCall(id="1", arguments=r'{"input": "task"}', name="_pass_function")]
+        assert isinstance(create_result.content, FunctionCalls)
+        assert create_result.content.function_calls == [
+            FunctionCall(id="1", arguments=r'{"input": "task"}', name="_pass_function")
+        ]
         assert create_result.finish_reason == "function_calls"
 
     # Warning completion when content is not None.
@@ -649,7 +656,10 @@ async def test_tool_calling(monkeypatch: pytest.MonkeyPatch) -> None:
         create_result = await model_client.create(
             messages=[UserMessage(content="Hello", source="user")], tools=[pass_tool]
         )
-        assert create_result.content == [FunctionCall(id="1", arguments=r'{"input": "task"}', name="_pass_function")]
+        assert isinstance(create_result.content, FunctionCalls)
+        assert create_result.content.function_calls == [
+            FunctionCall(id="1", arguments=r'{"input": "task"}', name="_pass_function")
+        ]
         assert create_result.finish_reason == "function_calls"
 
 
@@ -671,9 +681,10 @@ async def _test_model_client(model_client: OpenAIChatCompletionClient) -> None:
     create_result = await model_client.create(messages=messages, tools=[pass_tool, fail_tool])
     assert isinstance(create_result.content, list)
     assert len(create_result.content) == 1
-    assert isinstance(create_result.content[0], FunctionCall)
-    assert create_result.content[0].name == "pass_tool"
-    assert json.loads(create_result.content[0].arguments) == {"input": "task"}
+    assert isinstance(create_result.content, FunctionCalls)
+    assert isinstance(create_result.content.function_calls[0], FunctionCall)
+    assert create_result.content.function_calls[0].name == "pass_tool"
+    assert json.loads(create_result.content.function_calls[0].arguments) == {"input": "task"}
     assert create_result.finish_reason == "function_calls"
     assert create_result.usage is not None
 
@@ -681,7 +692,7 @@ async def _test_model_client(model_client: OpenAIChatCompletionClient) -> None:
     messages.append(AssistantMessage(content=create_result.content, source="assistant"))
     messages.append(
         FunctionExecutionResultMessage(
-            content=[FunctionExecutionResult(content="passed", call_id=create_result.content[0].id)]
+            content=[FunctionExecutionResult(content="passed", call_id=create_result.content.function_calls[0].id)]
         )
     )
     create_result = await model_client.create(messages=messages)
@@ -697,12 +708,13 @@ async def _test_model_client(model_client: OpenAIChatCompletionClient) -> None:
     create_result = await model_client.create(messages=messages, tools=[pass_tool, fail_tool])
     assert isinstance(create_result.content, list)
     assert len(create_result.content) == 2
-    assert isinstance(create_result.content[0], FunctionCall)
-    assert create_result.content[0].name == "pass_tool"
-    assert json.loads(create_result.content[0].arguments) == {"input": "task"}
+    assert isinstance(create_result.content, FunctionCalls)
+    assert isinstance(create_result.content.function_calls[0], FunctionCall)
+    assert create_result.content.function_calls[0].name == "pass_tool"
+    assert json.loads(create_result.content.function_calls[0].arguments) == {"input": "task"}
     assert isinstance(create_result.content[1], FunctionCall)
-    assert create_result.content[1].name == "fail_tool"
-    assert json.loads(create_result.content[1].arguments) == {"input": "task"}
+    assert create_result.content.function_calls[1].name == "fail_tool"
+    assert json.loads(create_result.content.function_calls[1].arguments) == {"input": "task"}
     assert create_result.finish_reason == "function_calls"
     assert create_result.usage is not None
 
@@ -711,8 +723,8 @@ async def _test_model_client(model_client: OpenAIChatCompletionClient) -> None:
     messages.append(
         FunctionExecutionResultMessage(
             content=[
-                FunctionExecutionResult(content="passed", call_id=create_result.content[0].id),
-                FunctionExecutionResult(content="failed", call_id=create_result.content[1].id),
+                FunctionExecutionResult(content="passed", call_id=create_result.content.function_calls[0].id),
+                FunctionExecutionResult(content="failed", call_id=create_result.content.function_calls[1].id),
             ]
         )
     )
diff --git a/python/packages/autogen-ext/tests/models/test_sk_chat_completion_adapter.py b/python/packages/autogen-ext/tests/models/test_sk_chat_completion_adapter.py
index 69be349c38c9..9edb25e185ec 100644
--- a/python/packages/autogen-ext/tests/models/test_sk_chat_completion_adapter.py
+++ b/python/packages/autogen-ext/tests/models/test_sk_chat_completion_adapter.py
@@ -3,7 +3,7 @@
 from unittest.mock import AsyncMock
 
 import pytest
-from autogen_core import CancellationToken
+from autogen_core import CancellationToken, FunctionCalls
 from autogen_core.models import CreateResult, LLMMessage, ModelFamily, ModelInfo, SystemMessage, UserMessage
 from autogen_core.tools import BaseTool
 from autogen_ext.models.semantic_kernel import SKChatCompletionAdapter
@@ -252,7 +252,7 @@ async def test_sk_chat_completion_with_tools(sk_client: AzureChatCompletion) ->
     result = await adapter.create(messages=messages, tools=[tool], extra_create_args={"kernel": kernel})
 
     # Verify response
-    assert isinstance(result.content, list)
+    assert isinstance(result.content, FunctionCalls)
     assert result.finish_reason == "function_calls"
     assert result.usage.prompt_tokens >= 0
     assert result.usage.completion_tokens >= 0
@@ -306,7 +306,7 @@ async def test_sk_chat_completion_stream_with_tools(sk_client: AzureChatCompleti
     assert len(response_chunks) > 0
     final_chunk = response_chunks[-1]
     assert isinstance(final_chunk, CreateResult)
-    assert isinstance(final_chunk.content, list)  # Function calls
+    assert isinstance(final_chunk.content, FunctionCalls)  # Function calls
     assert final_chunk.finish_reason == "function_calls"
     assert final_chunk.usage.prompt_tokens >= 0
     assert final_chunk.usage.completion_tokens >= 0
diff --git a/python/packages/autogen-ext/tests/test_filesurfer_agent.py b/python/packages/autogen-ext/tests/test_filesurfer_agent.py
index 513a1fe8444c..a48e03c5a594 100644
--- a/python/packages/autogen-ext/tests/test_filesurfer_agent.py
+++ b/python/packages/autogen-ext/tests/test_filesurfer_agent.py
@@ -140,8 +140,10 @@ async def test_run_filesurfer(monkeypatch: pytest.MonkeyPatch) -> None:
     # Get the FileSurfer to read the file, and the directory
     assert agent._name == "FileSurfer"  # pyright: ignore[reportPrivateUsage]
     result = await agent.run(task="Please read the test file")
+    assert isinstance(result.messages[1].content, str)
     assert "# FileSurfer test H1" in result.messages[1].content
 
     result = await agent.run(task="Please read the test directory")
+    assert isinstance(result.messages[1].content, str)
     assert "# Index of " in result.messages[1].content
     assert "test_filesurfer_agent.html" in result.messages[1].content
diff --git a/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/file_surfer/file_surfer.py b/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/file_surfer/file_surfer.py
index 7e17acdeb0f7..cc3dc6110d49 100644
--- a/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/file_surfer/file_surfer.py
+++ b/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/file_surfer/file_surfer.py
@@ -2,7 +2,7 @@
 import time
 from typing import List, Optional, Tuple
 
-from autogen_core import CancellationToken, FunctionCall, default_subscription
+from autogen_core import CancellationToken, FunctionCall, FunctionCalls, default_subscription
 from autogen_core.models import (
     ChatCompletionClient,
     SystemMessage,
@@ -99,8 +99,10 @@ async def _generate_reply(self, cancellation_token: CancellationToken) -> Tuple[
         if isinstance(response, str):
             return False, response
 
-        elif isinstance(response, list) and all(isinstance(item, FunctionCall) for item in response):
-            function_calls = response
+        elif isinstance(response, FunctionCalls) and all(
+            isinstance(item, FunctionCall) for item in response.function_calls
+        ):
+            function_calls = response.function_calls
             for function_call in function_calls:
                 tool_name = function_call.name
 
diff --git a/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/orchestrator.py b/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/orchestrator.py
index 2bcd3020fb64..0a89b6c8a2a2 100644
--- a/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/orchestrator.py
+++ b/python/packages/autogen-magentic-one/src/autogen_magentic_one/agents/orchestrator.py
@@ -1,7 +1,7 @@
 import json
 from typing import Any, Dict, List, Optional
 
-from autogen_core import AgentProxy, CancellationToken, MessageContext, TopicId, default_subscription
+from autogen_core import AgentProxy, CancellationToken, FunctionCalls, MessageContext, TopicId, default_subscription
 from autogen_core.models import (
     AssistantMessage,
     ChatCompletionClient,
@@ -130,9 +130,12 @@ def _get_message_str(self, message: LLMMessage) -> str:
             return message.content
         else:
             result = ""
-            for content in message.content:
-                if isinstance(content, str):
-                    result += content + "\n"
+            if isinstance(message.content, FunctionCalls) and message.content.thought:
+                return message.content.thought
+            elif isinstance(message.content, list):
+                for content in message.content:
+                    if isinstance(content, str):
+                        result += content + "\n"
             assert len(result) > 0
         return result
 
diff --git a/python/packages/autogen-magentic-one/src/autogen_magentic_one/messages.py b/python/packages/autogen-magentic-one/src/autogen_magentic_one/messages.py
index b4ffdd66b1ad..44d4ebb70dc4 100644
--- a/python/packages/autogen-magentic-one/src/autogen_magentic_one/messages.py
+++ b/python/packages/autogen-magentic-one/src/autogen_magentic_one/messages.py
@@ -1,13 +1,13 @@
 from dataclasses import dataclass
 from typing import Any, Dict, List, Union
 
-from autogen_core import FunctionCall, Image
+from autogen_core import FunctionCalls, Image
 from autogen_core.models import FunctionExecutionResult, LLMMessage
 from pydantic import BaseModel
 
 # Convenience type
 UserContent = Union[str, List[Union[str, Image]]]
-AssistantContent = Union[str, List[FunctionCall]]
+AssistantContent = Union[str, FunctionCalls]
 FunctionExecutionContent = List[FunctionExecutionResult]
 SystemContent = str
 
diff --git a/python/samples/core_async_human_in_the_loop/main.py b/python/samples/core_async_human_in_the_loop/main.py
index ccae231b2118..ce483f3d0cbf 100644
--- a/python/samples/core_async_human_in_the_loop/main.py
+++ b/python/samples/core_async_human_in_the_loop/main.py
@@ -35,6 +35,7 @@
     DefaultInterventionHandler,
     DefaultTopicId,
     FunctionCall,
+    FunctionCalls,
     MessageContext,
     RoutedAgent,
     SingleThreadedAgentRuntime,
@@ -182,8 +183,8 @@ async def handle_message(self, message: UserTextMessage, ctx: MessageContext) ->
             self._system_messages + (await self._model_context.get_messages()), tools=tools
         )
 
-        if isinstance(response.content, list) and all(isinstance(item, FunctionCall) for item in response.content):
-            for call in response.content:
+        if isinstance(response.content, FunctionCalls) and all(isinstance(item, FunctionCall) for item in response.content.function_calls):
+            for call in response.content.function_calls:
                 tool = next((tool for tool in tools if tool.name == call.name), None)
                 if tool is None:
                     raise ValueError(f"Tool not found: {call.name}")
diff --git a/python/uv.lock b/python/uv.lock
index 21adda9f946b..1803fd701f01 100644
--- a/python/uv.lock
+++ b/python/uv.lock
@@ -89,7 +89,6 @@ wheels = [
 
 [[package]]
 name = "agbench"
-version = "0.0.1a1"
 source = { editable = "packages/agbench" }
 dependencies = [
     { name = "azure-identity" },
@@ -796,7 +795,6 @@ requires-dist = [
 
 [[package]]
 name = "autogenstudio"
-version = "0.4.0"
 source = { editable = "packages/autogen-studio" }
 dependencies = [
     { name = "aiofiles" },
@@ -4217,7 +4215,6 @@ name = "nvidia-cublas-cu12"
 version = "12.4.5.8"
 source = { registry = "https://pypi.org/simple" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/7f/7f/7fbae15a3982dc9595e49ce0f19332423b260045d0a6afe93cdbe2f1f624/nvidia_cublas_cu12-12.4.5.8-py3-none-manylinux2014_aarch64.whl", hash = "sha256:0f8aa1706812e00b9f19dfe0cdb3999b092ccb8ca168c0db5b8ea712456fd9b3", size = 363333771 },
     { url = "https://files.pythonhosted.org/packages/ae/71/1c91302526c45ab494c23f61c7a84aa568b8c1f9d196efa5993957faf906/nvidia_cublas_cu12-12.4.5.8-py3-none-manylinux2014_x86_64.whl", hash = "sha256:2fc8da60df463fdefa81e323eef2e36489e1c94335b5358bcb38360adf75ac9b", size = 363438805 },
 ]
 
@@ -4226,7 +4223,6 @@ name = "nvidia-cuda-cupti-cu12"
 version = "12.4.127"
 source = { registry = "https://pypi.org/simple" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/93/b5/9fb3d00386d3361b03874246190dfec7b206fd74e6e287b26a8fcb359d95/nvidia_cuda_cupti_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:79279b35cf6f91da114182a5ce1864997fd52294a87a16179ce275773799458a", size = 12354556 },
     { url = "https://files.pythonhosted.org/packages/67/42/f4f60238e8194a3106d06a058d494b18e006c10bb2b915655bd9f6ea4cb1/nvidia_cuda_cupti_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:9dec60f5ac126f7bb551c055072b69d85392b13311fcc1bcda2202d172df30fb", size = 13813957 },
 ]
 
@@ -4235,7 +4231,6 @@ name = "nvidia-cuda-nvrtc-cu12"
 version = "12.4.127"
 source = { registry = "https://pypi.org/simple" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/77/aa/083b01c427e963ad0b314040565ea396f914349914c298556484f799e61b/nvidia_cuda_nvrtc_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:0eedf14185e04b76aa05b1fea04133e59f465b6f960c0cbf4e37c3cb6b0ea198", size = 24133372 },
     { url = "https://files.pythonhosted.org/packages/2c/14/91ae57cd4db3f9ef7aa99f4019cfa8d54cb4caa7e00975df6467e9725a9f/nvidia_cuda_nvrtc_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:a178759ebb095827bd30ef56598ec182b85547f1508941a3d560eb7ea1fbf338", size = 24640306 },
 ]
 
@@ -4244,7 +4239,6 @@ name = "nvidia-cuda-runtime-cu12"
 version = "12.4.127"
 source = { registry = "https://pypi.org/simple" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/a1/aa/b656d755f474e2084971e9a297def515938d56b466ab39624012070cb773/nvidia_cuda_runtime_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:961fe0e2e716a2a1d967aab7caee97512f71767f852f67432d572e36cb3a11f3", size = 894177 },
     { url = "https://files.pythonhosted.org/packages/ea/27/1795d86fe88ef397885f2e580ac37628ed058a92ed2c39dc8eac3adf0619/nvidia_cuda_runtime_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:64403288fa2136ee8e467cdc9c9427e0434110899d07c779f25b5c068934faa5", size = 883737 },
 ]
 
@@ -4267,7 +4261,6 @@ dependencies = [
     { name = "nvidia-nvjitlink-cu12", marker = "(platform_machine != 'aarch64' and sys_platform == 'linux') or (sys_platform != 'darwin' and sys_platform != 'linux')" },
 ]
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/7a/8a/0e728f749baca3fbeffad762738276e5df60851958be7783af121a7221e7/nvidia_cufft_cu12-11.2.1.3-py3-none-manylinux2014_aarch64.whl", hash = "sha256:5dad8008fc7f92f5ddfa2101430917ce2ffacd86824914c82e28990ad7f00399", size = 211422548 },
     { url = "https://files.pythonhosted.org/packages/27/94/3266821f65b92b3138631e9c8e7fe1fb513804ac934485a8d05776e1dd43/nvidia_cufft_cu12-11.2.1.3-py3-none-manylinux2014_x86_64.whl", hash = "sha256:f083fc24912aa410be21fa16d157fed2055dab1cc4b6934a0e03cba69eb242b9", size = 211459117 },
 ]
 
@@ -4276,7 +4269,6 @@ name = "nvidia-curand-cu12"
 version = "10.3.5.147"
 source = { registry = "https://pypi.org/simple" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/80/9c/a79180e4d70995fdf030c6946991d0171555c6edf95c265c6b2bf7011112/nvidia_curand_cu12-10.3.5.147-py3-none-manylinux2014_aarch64.whl", hash = "sha256:1f173f09e3e3c76ab084aba0de819c49e56614feae5c12f69883f4ae9bb5fad9", size = 56314811 },
     { url = "https://files.pythonhosted.org/packages/8a/6d/44ad094874c6f1b9c654f8ed939590bdc408349f137f9b98a3a23ccec411/nvidia_curand_cu12-10.3.5.147-py3-none-manylinux2014_x86_64.whl", hash = "sha256:a88f583d4e0bb643c49743469964103aa59f7f708d862c3ddb0fc07f851e3b8b", size = 56305206 },
 ]
 
@@ -4290,7 +4282,6 @@ dependencies = [
     { name = "nvidia-nvjitlink-cu12", marker = "(platform_machine != 'aarch64' and sys_platform == 'linux') or (sys_platform != 'darwin' and sys_platform != 'linux')" },
 ]
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/46/6b/a5c33cf16af09166845345275c34ad2190944bcc6026797a39f8e0a282e0/nvidia_cusolver_cu12-11.6.1.9-py3-none-manylinux2014_aarch64.whl", hash = "sha256:d338f155f174f90724bbde3758b7ac375a70ce8e706d70b018dd3375545fc84e", size = 127634111 },
     { url = "https://files.pythonhosted.org/packages/3a/e1/5b9089a4b2a4790dfdea8b3a006052cfecff58139d5a4e34cb1a51df8d6f/nvidia_cusolver_cu12-11.6.1.9-py3-none-manylinux2014_x86_64.whl", hash = "sha256:19e33fa442bcfd085b3086c4ebf7e8debc07cfe01e11513cc6d332fd918ac260", size = 127936057 },
 ]
 
@@ -4302,7 +4293,6 @@ dependencies = [
     { name = "nvidia-nvjitlink-cu12", marker = "(platform_machine != 'aarch64' and sys_platform == 'linux') or (sys_platform != 'darwin' and sys_platform != 'linux')" },
 ]
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/96/a9/c0d2f83a53d40a4a41be14cea6a0bf9e668ffcf8b004bd65633f433050c0/nvidia_cusparse_cu12-12.3.1.170-py3-none-manylinux2014_aarch64.whl", hash = "sha256:9d32f62896231ebe0480efd8a7f702e143c98cfaa0e8a76df3386c1ba2b54df3", size = 207381987 },
     { url = "https://files.pythonhosted.org/packages/db/f7/97a9ea26ed4bbbfc2d470994b8b4f338ef663be97b8f677519ac195e113d/nvidia_cusparse_cu12-12.3.1.170-py3-none-manylinux2014_x86_64.whl", hash = "sha256:ea4f11a2904e2a8dc4b1833cc1b5181cde564edd0d5cd33e3c168eff2d1863f1", size = 207454763 },
 ]
 
@@ -4319,7 +4309,6 @@ name = "nvidia-nvjitlink-cu12"
 version = "12.4.127"
 source = { registry = "https://pypi.org/simple" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/02/45/239d52c05074898a80a900f49b1615d81c07fceadd5ad6c4f86a987c0bc4/nvidia_nvjitlink_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:4abe7fef64914ccfa909bc2ba39739670ecc9e820c83ccc7a6ed414122599b83", size = 20552510 },
     { url = "https://files.pythonhosted.org/packages/ff/ff/847841bacfbefc97a00036e0fce5a0f086b640756dc38caea5e1bb002655/nvidia_nvjitlink_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:06b3b9b25bf3f8af351d664978ca26a16d2c5127dbd53c0497e28d1fb9611d57", size = 21066810 },
 ]
 
@@ -4328,7 +4317,6 @@ name = "nvidia-nvtx-cu12"
 version = "12.4.127"
 source = { registry = "https://pypi.org/simple" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/06/39/471f581edbb7804b39e8063d92fc8305bdc7a80ae5c07dbe6ea5c50d14a5/nvidia_nvtx_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:7959ad635db13edf4fc65c06a6e9f9e55fc2f92596db928d169c0bb031e88ef3", size = 100417 },
     { url = "https://files.pythonhosted.org/packages/87/20/199b8713428322a2f22b722c62b8cc278cc53dffa9705d744484b5035ee9/nvidia_nvtx_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:781e950d9b9f60d8241ccea575b32f5105a5baf4c2351cab5256a24869f12a1a", size = 99144 },
 ]