pydantic-ai-slim 1.9.0__tar.gz → 1.10.0__tar.gz

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pydantic-ai-slim
-Version: 1.9.0
+Version: 1.10.0
 Summary: Agent Framework / shim to use Pydantic with LLMs, slim package
 Project-URL: Homepage, https://github.com/pydantic/pydantic-ai/tree/main/pydantic_ai_slim
 Project-URL: Source, https://github.com/pydantic/pydantic-ai/tree/main/pydantic_ai_slim
@@ -33,7 +33,7 @@ Requires-Dist: genai-prices>=0.0.35
 Requires-Dist: griffe>=1.3.2
 Requires-Dist: httpx>=0.27
 Requires-Dist: opentelemetry-api>=1.28.0
-Requires-Dist: pydantic-graph==1.9.0
+Requires-Dist: pydantic-graph==1.10.0
 Requires-Dist: pydantic>=2.10
 Requires-Dist: typing-inspection>=0.4.0
 Provides-Extra: a2a
@@ -57,7 +57,7 @@ Requires-Dist: dbos>=1.14.0; extra == 'dbos'
 Provides-Extra: duckduckgo
 Requires-Dist: ddgs>=9.0.0; extra == 'duckduckgo'
 Provides-Extra: evals
-Requires-Dist: pydantic-evals==1.9.0; extra == 'evals'
+Requires-Dist: pydantic-evals==1.10.0; extra == 'evals'
 Provides-Extra: fastmcp
 Requires-Dist: fastmcp>=2.12.0; extra == 'fastmcp'
 Provides-Extra: google

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/_run_context.py

@@ -16,15 +16,19 @@ if TYPE_CHECKING:
     from .models import Model
     from .result import RunUsage
 
+# TODO (v2): Change the default for all typevars like this from `None` to `object`
 AgentDepsT = TypeVar('AgentDepsT', default=None, contravariant=True)
 """Type variable for agent dependencies."""
 
+RunContextAgentDepsT = TypeVar('RunContextAgentDepsT', default=None, covariant=True)
+"""Type variable for the agent dependencies in `RunContext`."""
+
 
 @dataclasses.dataclass(repr=False, kw_only=True)
-class RunContext(Generic[AgentDepsT]):
+class RunContext(Generic[RunContextAgentDepsT]):
     """Information about the current call."""
 
-    deps: AgentDepsT
+    deps: RunContextAgentDepsT
     """Dependencies for the agent."""
     model: Model
     """The model used in this run."""

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/_utils.py

@@ -234,6 +234,15 @@ def sync_anext(iterator: Iterator[T]) -> T:
         raise StopAsyncIteration() from e
 
 
+def sync_async_iterator(async_iter: AsyncIterator[T]) -> Iterator[T]:
+    loop = get_event_loop()
+    while True:
+        try:
+            yield loop.run_until_complete(anext(async_iter))
+        except StopAsyncIteration:
+            break
+
+
 def now_utc() -> datetime:
     return datetime.now(tz=timezone.utc)
 
@@ -489,3 +498,12 @@ def get_union_args(tp: Any) -> tuple[Any, ...]:
         return tuple(_unwrap_annotated(arg) for arg in get_args(tp))
     else:
         return ()
+
+
+def get_event_loop():
+    try:
+        event_loop = asyncio.get_event_loop()
+    except RuntimeError:  # pragma: lax no cover
+        event_loop = asyncio.new_event_loop()
+        asyncio.set_event_loop(event_loop)
+    return event_loop

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/agent/abstract.py

@@ -12,7 +12,6 @@ import anyio
 from typing_extensions import Self, TypeIs, TypeVar
 
 from pydantic_graph import End
-from pydantic_graph._utils import get_event_loop
 
 from .. import (
     _agent_graph,
@@ -335,7 +334,7 @@ class AbstractAgent(Generic[AgentDepsT, OutputDataT], ABC):
         if infer_name and self.name is None:
             self._infer_name(inspect.currentframe())
 
-        return get_event_loop().run_until_complete(
+        return _utils.get_event_loop().run_until_complete(
             self.run(
                 user_prompt,
                 output_type=output_type,
@@ -581,6 +580,133 @@ class AbstractAgent(Generic[AgentDepsT, OutputDataT], ABC):
         if not yielded:
             raise exceptions.AgentRunError('Agent run finished without producing a final result')  # pragma: no cover
 
+    @overload
+    def run_stream_sync(
+        self,
+        user_prompt: str | Sequence[_messages.UserContent] | None = None,
+        *,
+        output_type: None = None,
+        message_history: Sequence[_messages.ModelMessage] | None = None,
+        deferred_tool_results: DeferredToolResults | None = None,
+        model: models.Model | models.KnownModelName | str | None = None,
+        deps: AgentDepsT = None,
+        model_settings: ModelSettings | None = None,
+        usage_limits: _usage.UsageLimits | None = None,
+        usage: _usage.RunUsage | None = None,
+        infer_name: bool = True,
+        toolsets: Sequence[AbstractToolset[AgentDepsT]] | None = None,
+        builtin_tools: Sequence[AbstractBuiltinTool] | None = None,
+        event_stream_handler: EventStreamHandler[AgentDepsT] | None = None,
+    ) -> result.StreamedRunResultSync[AgentDepsT, OutputDataT]: ...
+
+    @overload
+    def run_stream_sync(
+        self,
+        user_prompt: str | Sequence[_messages.UserContent] | None = None,
+        *,
+        output_type: OutputSpec[RunOutputDataT],
+        message_history: Sequence[_messages.ModelMessage] | None = None,
+        deferred_tool_results: DeferredToolResults | None = None,
+        model: models.Model | models.KnownModelName | str | None = None,
+        deps: AgentDepsT = None,
+        model_settings: ModelSettings | None = None,
+        usage_limits: _usage.UsageLimits | None = None,
+        usage: _usage.RunUsage | None = None,
+        infer_name: bool = True,
+        toolsets: Sequence[AbstractToolset[AgentDepsT]] | None = None,
+        builtin_tools: Sequence[AbstractBuiltinTool] | None = None,
+        event_stream_handler: EventStreamHandler[AgentDepsT] | None = None,
+    ) -> result.StreamedRunResultSync[AgentDepsT, RunOutputDataT]: ...
+
+    def run_stream_sync(
+        self,
+        user_prompt: str | Sequence[_messages.UserContent] | None = None,
+        *,
+        output_type: OutputSpec[RunOutputDataT] | None = None,
+        message_history: Sequence[_messages.ModelMessage] | None = None,
+        deferred_tool_results: DeferredToolResults | None = None,
+        model: models.Model | models.KnownModelName | str | None = None,
+        deps: AgentDepsT = None,
+        model_settings: ModelSettings | None = None,
+        usage_limits: _usage.UsageLimits | None = None,
+        usage: _usage.RunUsage | None = None,
+        infer_name: bool = True,
+        toolsets: Sequence[AbstractToolset[AgentDepsT]] | None = None,
+        builtin_tools: Sequence[AbstractBuiltinTool] | None = None,
+        event_stream_handler: EventStreamHandler[AgentDepsT] | None = None,
+    ) -> result.StreamedRunResultSync[AgentDepsT, Any]:
+        """Run the agent with a user prompt in sync streaming mode.
+
+        This is a convenience method that wraps [`run_stream()`][pydantic_ai.agent.AbstractAgent.run_stream] with `loop.run_until_complete(...)`.
+        You therefore can't use this method inside async code or if there's an active event loop.
+
+        This method builds an internal agent graph (using system prompts, tools and output schemas) and then
+        runs the graph until the model produces output matching the `output_type`, for example text or structured data.
+        At this point, a streaming run result object is yielded from which you can stream the output as it comes in,
+        and -- once this output has completed streaming -- get the complete output, message history, and usage.
+
+        As this method will consider the first output matching the `output_type` to be the final output,
+        it will stop running the agent graph and will not execute any tool calls made by the model after this "final" output.
+        If you want to always run the agent graph to completion and stream events and output at the same time,
+        use [`agent.run()`][pydantic_ai.agent.AbstractAgent.run] with an `event_stream_handler` or [`agent.iter()`][pydantic_ai.agent.AbstractAgent.iter] instead.
+
+        Example:
+        ```python
+        from pydantic_ai import Agent
+
+        agent = Agent('openai:gpt-4o')
+
+        def main():
+            response = agent.run_stream_sync('What is the capital of the UK?')
+            print(response.get_output())
+            #> The capital of the UK is London.
+        ```
+
+        Args:
+            user_prompt: User input to start/continue the conversation.
+            output_type: Custom output type to use for this run, `output_type` may only be used if the agent has no
+                output validators since output validators would expect an argument that matches the agent's output type.
+            message_history: History of the conversation so far.
+            deferred_tool_results: Optional results for deferred tool calls in the message history.
+            model: Optional model to use for this run, required if `model` was not set when creating the agent.
+            deps: Optional dependencies to use for this run.
+            model_settings: Optional settings to use for this model's request.
+            usage_limits: Optional limits on model request count or token usage.
+            usage: Optional usage to start with, useful for resuming a conversation or agents used in tools.
+            infer_name: Whether to try to infer the agent name from the call frame if it's not set.
+            toolsets: Optional additional toolsets for this run.
+            builtin_tools: Optional additional builtin tools for this run.
+            event_stream_handler: Optional handler for events from the model's streaming response and the agent's execution of tools to use for this run.
+                It will receive all the events up until the final result is found, which you can then read or stream from inside the context manager.
+                Note that it does _not_ receive any events after the final result is found.
+
+        Returns:
+            The result of the run.
+        """
+        if infer_name and self.name is None:
+            self._infer_name(inspect.currentframe())
+
+        async def _consume_stream():
+            async with self.run_stream(
+                user_prompt,
+                output_type=output_type,
+                message_history=message_history,
+                deferred_tool_results=deferred_tool_results,
+                model=model,
+                deps=deps,
+                model_settings=model_settings,
+                usage_limits=usage_limits,
+                usage=usage,
+                infer_name=infer_name,
+                toolsets=toolsets,
+                builtin_tools=builtin_tools,
+                event_stream_handler=event_stream_handler,
+            ) as stream_result:
+                yield stream_result
+
+        async_result = _utils.get_event_loop().run_until_complete(anext(_consume_stream()))
+        return result.StreamedRunResultSync(async_result)
+
     @overload
     def run_stream_events(
         self,
@@ -1217,6 +1343,6 @@ class AbstractAgent(Generic[AgentDepsT, OutputDataT], ABC):
         agent.to_cli_sync(prog_name='assistant')
         ```
         """
-        return get_event_loop().run_until_complete(
+        return _utils.get_event_loop().run_until_complete(
             self.to_cli(deps=deps, prog_name=prog_name, message_history=message_history)
         )

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/durable_exec/temporal/_function_toolset.py

@@ -1,57 +1,22 @@
 from __future__ import annotations
 
 from collections.abc import Callable
-from dataclasses import dataclass
-from typing import Annotated, Any, Literal
+from typing import Any, Literal
 
-from pydantic import ConfigDict, Discriminator, with_config
 from temporalio import activity, workflow
 from temporalio.workflow import ActivityConfig
-from typing_extensions import assert_never
 
 from pydantic_ai import FunctionToolset, ToolsetTool
-from pydantic_ai.exceptions import ApprovalRequired, CallDeferred, ModelRetry, UserError
+from pydantic_ai.exceptions import UserError
 from pydantic_ai.tools import AgentDepsT, RunContext
 from pydantic_ai.toolsets.function import FunctionToolsetTool
 
 from ._run_context import TemporalRunContext
-from ._toolset import TemporalWrapperToolset
-
-
-@dataclass
-@with_config(ConfigDict(arbitrary_types_allowed=True))
-class _CallToolParams:
-    name: str
-    tool_args: dict[str, Any]
-    serialized_run_context: Any
-
-
-@dataclass
-class _ApprovalRequired:
-    kind: Literal['approval_required'] = 'approval_required'
-
-
-@dataclass
-class _CallDeferred:
-    kind: Literal['call_deferred'] = 'call_deferred'
-
-
-@dataclass
-class _ModelRetry:
-    message: str
-    kind: Literal['model_retry'] = 'model_retry'
-
-
-@dataclass
-class _ToolReturn:
-    result: Any
-    kind: Literal['tool_return'] = 'tool_return'
-
-
-_CallToolResult = Annotated[
-    _ApprovalRequired | _CallDeferred | _ModelRetry | _ToolReturn,
-    Discriminator('kind'),
-]
+from ._toolset import (
+    CallToolParams,
+    CallToolResult,
+    TemporalWrapperToolset,
+)
 
 
 class TemporalFunctionToolset(TemporalWrapperToolset[AgentDepsT]):
@@ -70,7 +35,7 @@ class TemporalFunctionToolset(TemporalWrapperToolset[AgentDepsT]):
         self.tool_activity_config = tool_activity_config
         self.run_context_type = run_context_type
 
-        async def call_tool_activity(params: _CallToolParams, deps: AgentDepsT) -> _CallToolResult:
+        async def call_tool_activity(params: CallToolParams, deps: AgentDepsT) -> CallToolResult:
             name = params.name
             ctx = self.run_context_type.deserialize_run_context(params.serialized_run_context, deps=deps)
             try:
@@ -84,15 +49,7 @@ class TemporalFunctionToolset(TemporalWrapperToolset[AgentDepsT]):
             # The tool args will already have been validated into their proper types in the `ToolManager`,
             # but `execute_activity` would have turned them into simple Python types again, so we need to re-validate them.
             args_dict = tool.args_validator.validate_python(params.tool_args)
-            try:
-                result = await self.wrapped.call_tool(name, args_dict, ctx, tool)
-                return _ToolReturn(result=result)
-            except ApprovalRequired:
-                return _ApprovalRequired()
-            except CallDeferred:
-                return _CallDeferred()
-            except ModelRetry as e:
-                return _ModelRetry(message=e.message)
+            return await self._wrap_call_tool_result(self.wrapped.call_tool(name, args_dict, ctx, tool))
 
         # Set type hint explicitly so that Temporal can take care of serialization and deserialization
         call_tool_activity.__annotations__['deps'] = deps_type
@@ -123,25 +80,18 @@ class TemporalFunctionToolset(TemporalWrapperToolset[AgentDepsT]):
 
         tool_activity_config = self.activity_config | tool_activity_config
         serialized_run_context = self.run_context_type.serialize_run_context(ctx)
-        result = await workflow.execute_activity(  # pyright: ignore[reportUnknownMemberType]
-            activity=self.call_tool_activity,
-            args=[
-                _CallToolParams(
-                    name=name,
-                    tool_args=tool_args,
-                    serialized_run_context=serialized_run_context,
-                ),
-                ctx.deps,
-            ],
-            **tool_activity_config,
+        return self._unwrap_call_tool_result(
+            await workflow.execute_activity(  # pyright: ignore[reportUnknownMemberType]
+                activity=self.call_tool_activity,
+                args=[
+                    CallToolParams(
+                        name=name,
+                        tool_args=tool_args,
+                        serialized_run_context=serialized_run_context,
+                        tool_def=None,
+                    ),
+                    ctx.deps,
+                ],
+                **tool_activity_config,
+            )
         )
-        if isinstance(result, _ApprovalRequired):
-            raise ApprovalRequired()
-        elif isinstance(result, _CallDeferred):
-            raise CallDeferred()
-        elif isinstance(result, _ModelRetry):
-            raise ModelRetry(result.message)
-        elif isinstance(result, _ToolReturn):
-            return result.result
-        else:
-            assert_never(result)

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/durable_exec/temporal/_mcp_server.py

@@ -11,11 +11,15 @@ from typing_extensions import Self
 
 from pydantic_ai import ToolsetTool
 from pydantic_ai.exceptions import UserError
-from pydantic_ai.mcp import MCPServer, ToolResult
+from pydantic_ai.mcp import MCPServer
 from pydantic_ai.tools import AgentDepsT, RunContext, ToolDefinition
 
 from ._run_context import TemporalRunContext
-from ._toolset import TemporalWrapperToolset
+from ._toolset import (
+    CallToolParams,
+    CallToolResult,
+    TemporalWrapperToolset,
+)
 
 
 @dataclass
@@ -24,15 +28,6 @@ class _GetToolsParams:
     serialized_run_context: Any
 
 
-@dataclass
-@with_config(ConfigDict(arbitrary_types_allowed=True))
-class _CallToolParams:
-    name: str
-    tool_args: dict[str, Any]
-    serialized_run_context: Any
-    tool_def: ToolDefinition
-
-
 class TemporalMCPServer(TemporalWrapperToolset[AgentDepsT]):
     def __init__(
         self,
@@ -72,13 +67,16 @@ class TemporalMCPServer(TemporalWrapperToolset[AgentDepsT]):
             get_tools_activity
         )
 
-        async def call_tool_activity(params: _CallToolParams, deps: AgentDepsT) -> ToolResult:
+        async def call_tool_activity(params: CallToolParams, deps: AgentDepsT) -> CallToolResult:
             run_context = self.run_context_type.deserialize_run_context(params.serialized_run_context, deps=deps)
-            return await self.wrapped.call_tool(
-                params.name,
-                params.tool_args,
-                run_context,
-                self.tool_for_tool_def(params.tool_def),
+            assert isinstance(params.tool_def, ToolDefinition)
+            return await self._wrap_call_tool_result(
+                self.wrapped.call_tool(
+                    params.name,
+                    params.tool_args,
+                    run_context,
+                    self.tool_for_tool_def(params.tool_def),
+                )
             )
 
         # Set type hint explicitly so that Temporal can take care of serialization and deserialization
@@ -125,22 +123,24 @@ class TemporalMCPServer(TemporalWrapperToolset[AgentDepsT]):
         tool_args: dict[str, Any],
         ctx: RunContext[AgentDepsT],
         tool: ToolsetTool[AgentDepsT],
-    ) -> ToolResult:
+    ) -> CallToolResult:
         if not workflow.in_workflow():
             return await super().call_tool(name, tool_args, ctx, tool)
 
         tool_activity_config = self.activity_config | self.tool_activity_config.get(name, {})
         serialized_run_context = self.run_context_type.serialize_run_context(ctx)
-        return await workflow.execute_activity(  # pyright: ignore[reportUnknownMemberType]
-            activity=self.call_tool_activity,
-            args=[
-                _CallToolParams(
-                    name=name,
-                    tool_args=tool_args,
-                    serialized_run_context=serialized_run_context,
-                    tool_def=tool.tool_def,
-                ),
-                ctx.deps,
-            ],
-            **tool_activity_config,
+        return self._unwrap_call_tool_result(
+            await workflow.execute_activity(  # pyright: ignore[reportUnknownMemberType]
+                activity=self.call_tool_activity,
+                args=[
+                    CallToolParams(
+                        name=name,
+                        tool_args=tool_args,
+                        serialized_run_context=serialized_run_context,
+                        tool_def=tool.tool_def,
+                    ),
+                    ctx.deps,
+                ],
+                **tool_activity_config,
+            )
         )

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/durable_exec/temporal/_run_context.py

@@ -2,8 +2,13 @@ from __future__ import annotations
 
 from typing import Any
 
+from typing_extensions import TypeVar
+
 from pydantic_ai.exceptions import UserError
-from pydantic_ai.tools import AgentDepsT, RunContext
+from pydantic_ai.tools import RunContext
+
+AgentDepsT = TypeVar('AgentDepsT', default=None, covariant=True)
+"""Type variable for the agent dependencies in `RunContext`."""
 
 
 class TemporalRunContext(RunContext[AgentDepsT]):
@@ -47,6 +52,6 @@ class TemporalRunContext(RunContext[AgentDepsT]):
         }
 
     @classmethod
-    def deserialize_run_context(cls, ctx: dict[str, Any], deps: AgentDepsT) -> TemporalRunContext[AgentDepsT]:
+    def deserialize_run_context(cls, ctx: dict[str, Any], deps: Any) -> TemporalRunContext[Any]:
         """Deserialize the run context from a `dict[str, Any]`."""
         return cls(**ctx, deps=deps)

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/durable_exec/temporal/_toolset.py

@@ -1,17 +1,58 @@
 from __future__ import annotations
 
 from abc import ABC, abstractmethod
-from collections.abc import Callable
-from typing import Any, Literal
+from collections.abc import Awaitable, Callable
+from dataclasses import dataclass
+from typing import Annotated, Any, Literal
 
+from pydantic import ConfigDict, Discriminator, with_config
 from temporalio.workflow import ActivityConfig
+from typing_extensions import assert_never
 
 from pydantic_ai import AbstractToolset, FunctionToolset, WrapperToolset
-from pydantic_ai.tools import AgentDepsT
+from pydantic_ai.exceptions import ApprovalRequired, CallDeferred, ModelRetry
+from pydantic_ai.tools import AgentDepsT, ToolDefinition
 
 from ._run_context import TemporalRunContext
 
 
+@dataclass
+@with_config(ConfigDict(arbitrary_types_allowed=True))
+class CallToolParams:
+    name: str
+    tool_args: dict[str, Any]
+    serialized_run_context: Any
+    tool_def: ToolDefinition | None
+
+
+@dataclass
+class _ApprovalRequired:
+    kind: Literal['approval_required'] = 'approval_required'
+
+
+@dataclass
+class _CallDeferred:
+    kind: Literal['call_deferred'] = 'call_deferred'
+
+
+@dataclass
+class _ModelRetry:
+    message: str
+    kind: Literal['model_retry'] = 'model_retry'
+
+
+@dataclass
+class _ToolReturn:
+    result: Any
+    kind: Literal['tool_return'] = 'tool_return'
+
+
+CallToolResult = Annotated[
+    _ApprovalRequired | _CallDeferred | _ModelRetry | _ToolReturn,
+    Discriminator('kind'),
+]
+
+
 class TemporalWrapperToolset(WrapperToolset[AgentDepsT], ABC):
     @property
     def id(self) -> str:
@@ -30,6 +71,29 @@ class TemporalWrapperToolset(WrapperToolset[AgentDepsT], ABC):
         # Temporalized toolsets cannot be swapped out after the fact.
         return self
 
+    async def _wrap_call_tool_result(self, coro: Awaitable[Any]) -> CallToolResult:
+        try:
+            result = await coro
+            return _ToolReturn(result=result)
+        except ApprovalRequired:
+            return _ApprovalRequired()
+        except CallDeferred:
+            return _CallDeferred()
+        except ModelRetry as e:
+            return _ModelRetry(message=e.message)
+
+    def _unwrap_call_tool_result(self, result: CallToolResult) -> Any:
+        if isinstance(result, _ToolReturn):
+            return result.result
+        elif isinstance(result, _ApprovalRequired):
+            raise ApprovalRequired()
+        elif isinstance(result, _CallDeferred):
+            raise CallDeferred()
+        elif isinstance(result, _ModelRetry):
+            raise ModelRetry(result.message)
+        else:
+            assert_never(result)
+
 
 def temporalize_toolset(
     toolset: AbstractToolset[AgentDepsT],

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/messages.py

@@ -34,6 +34,7 @@ DocumentMediaType: TypeAlias = Literal[
     'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
     'text/html',
     'text/markdown',
+    'application/msword',
     'application/vnd.ms-excel',
 ]
 VideoMediaType: TypeAlias = Literal[
@@ -434,8 +435,12 @@ class DocumentUrl(FileUrl):
             return 'application/pdf'
         elif self.url.endswith('.rtf'):
             return 'application/rtf'
+        elif self.url.endswith('.doc'):
+            return 'application/msword'
         elif self.url.endswith('.docx'):
             return 'application/vnd.openxmlformats-officedocument.wordprocessingml.document'
+        elif self.url.endswith('.xls'):
+            return 'application/vnd.ms-excel'
         elif self.url.endswith('.xlsx'):
             return 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet'
 
@@ -645,6 +650,7 @@ _document_format_lookup: dict[str, DocumentFormat] = {
     'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet': 'xlsx',
     'text/html': 'html',
     'text/markdown': 'md',
+    'application/msword': 'doc',
     'application/vnd.ms-excel': 'xls',
 }
 _audio_format_lookup: dict[str, AudioFormat] = {

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/models/google.py

@@ -471,11 +471,9 @@ class GoogleModel(Model):
                 raise UnexpectedModelBehavior(
                     f'Content filter {raw_finish_reason.value!r} triggered', response.model_dump_json()
                 )
-            else:
-                raise UnexpectedModelBehavior(
-                    'Content field missing from Gemini response', response.model_dump_json()
-                )  # pragma: no cover
-        parts = candidate.content.parts or []
+            parts = []  # pragma: no cover
+        else:
+            parts = candidate.content.parts or []
 
         usage = _metadata_as_usage(response)
         return _process_response_from_parts(
@@ -649,17 +647,12 @@ class GeminiStreamedResponse(StreamedResponse):
             #     )
 
             if candidate.content is None or candidate.content.parts is None:
-                if self.finish_reason == 'stop':  # pragma: no cover
-                    # Normal completion - skip this chunk
-                    continue
-                elif self.finish_reason == 'content_filter' and raw_finish_reason:  # pragma: no cover
+                if self.finish_reason == 'content_filter' and raw_finish_reason:  # pragma: no cover
                     raise UnexpectedModelBehavior(
                         f'Content filter {raw_finish_reason.value!r} triggered', chunk.model_dump_json()
                     )
                 else:  # pragma: no cover
-                    raise UnexpectedModelBehavior(
-                        'Content field missing from streaming Gemini response', chunk.model_dump_json()
-                    )
+                    continue
 
             parts = candidate.content.parts
             if not parts:

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/models/openai.py

@@ -948,6 +948,10 @@ class OpenAIResponsesModel(Model):
 
         super().__init__(settings=settings, profile=profile or provider.model_profile)
 
+    @property
+    def base_url(self) -> str:
+        return str(self.client.base_url)
+
     @property
     def model_name(self) -> OpenAIModelName:
         """The model name."""

{pydantic_ai_slim-1.9.0 → pydantic_ai_slim-1.10.0}/pydantic_ai/providers/anthropic.py

@@ -12,7 +12,7 @@ from pydantic_ai.profiles.anthropic import anthropic_model_profile
 from pydantic_ai.providers import Provider
 
 try:
-    from anthropic import AsyncAnthropic, AsyncAnthropicBedrock
+    from anthropic import AsyncAnthropic, AsyncAnthropicBedrock, AsyncAnthropicVertex
 except ImportError as _import_error:
     raise ImportError(
         'Please install the `anthropic` package to use the Anthropic provider, '
@@ -20,7 +20,7 @@ except ImportError as _import_error:
     ) from _import_error
 
 
-AsyncAnthropicClient: TypeAlias = AsyncAnthropic | AsyncAnthropicBedrock
+AsyncAnthropicClient: TypeAlias = AsyncAnthropic | AsyncAnthropicBedrock | AsyncAnthropicVertex
 
 
 class AnthropicProvider(Provider[AsyncAnthropicClient]):