pydantic-ai-slim 1.0.14__py3-none-any.whl → 1.0.15__py3-none-any.whl

pydantic_ai/durable_exec/temporal/_agent.py

@@ -17,6 +17,7 @@ from typing_extensions import Never
 
 from pydantic_ai import (
     AbstractToolset,
+    AgentRunResultEvent,
     _utils,
     messages as _messages,
     models,
@@ -558,9 +559,8 @@ class TemporalAgent(WrapperAgent[AgentDepsT, OutputDataT]):
         """
         if workflow.in_workflow():
             raise UserError(
-                '`agent.run_stream()` cannot currently be used inside a Temporal workflow. '
-                'Set an `event_stream_handler` on the agent and use `agent.run()` instead. '
-                'Please file an issue if this is not sufficient for your use case.'
+                '`agent.run_stream()` cannot be used inside a Temporal workflow. '
+                'Set an `event_stream_handler` on the agent and use `agent.run()` instead.'
             )
 
         async with super().run_stream(
@@ -580,6 +580,124 @@ class TemporalAgent(WrapperAgent[AgentDepsT, OutputDataT]):
         ) as result:
             yield result
 
+    @overload
+    def run_stream_events(
+        self,
+        user_prompt: str | Sequence[_messages.UserContent] | None = None,
+        *,
+        output_type: None = None,
+        message_history: list[_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,
+    ) -> AsyncIterator[_messages.AgentStreamEvent | AgentRunResultEvent[OutputDataT]]: ...
+
+    @overload
+    def run_stream_events(
+        self,
+        user_prompt: str | Sequence[_messages.UserContent] | None = None,
+        *,
+        output_type: OutputSpec[RunOutputDataT],
+        message_history: list[_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,
+    ) -> AsyncIterator[_messages.AgentStreamEvent | AgentRunResultEvent[RunOutputDataT]]: ...
+
+    def run_stream_events(
+        self,
+        user_prompt: str | Sequence[_messages.UserContent] | None = None,
+        *,
+        output_type: OutputSpec[RunOutputDataT] | None = None,
+        message_history: list[_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,
+    ) -> AsyncIterator[_messages.AgentStreamEvent | AgentRunResultEvent[Any]]:
+        """Run the agent with a user prompt in async mode and stream events from the run.
+
+        This is a convenience method that wraps [`self.run`][pydantic_ai.agent.AbstractAgent.run] and
+        uses the `event_stream_handler` kwarg to get a stream of events from the run.
+
+        Example:
+        ```python
+        from pydantic_ai import Agent, AgentRunResultEvent, AgentStreamEvent
+
+        agent = Agent('openai:gpt-4o')
+
+        async def main():
+            events: list[AgentStreamEvent | AgentRunResultEvent] = []
+            async for event in agent.run_stream_events('What is the capital of France?'):
+                events.append(event)
+            print(events)
+            '''
+            [
+                PartStartEvent(index=0, part=TextPart(content='The capital of ')),
+                FinalResultEvent(tool_name=None, tool_call_id=None),
+                PartDeltaEvent(index=0, delta=TextPartDelta(content_delta='France is Paris. ')),
+                AgentRunResultEvent(
+                    result=AgentRunResult(output='The capital of France is Paris. ')
+                ),
+            ]
+            '''
+        ```
+
+        Arguments are the same as for [`self.run`][pydantic_ai.agent.AbstractAgent.run],
+        except that `event_stream_handler` is now allowed.
+
+        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.
+
+        Returns:
+            An async iterable of stream events `AgentStreamEvent` and finally a `AgentRunResultEvent` with the final
+            run result.
+        """
+        if workflow.in_workflow():
+            raise UserError(
+                '`agent.run_stream_events()` cannot be used inside a Temporal workflow. '
+                'Set an `event_stream_handler` on the agent and use `agent.run()` instead.'
+            )
+
+        return super().run_stream_events(
+            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,
+        )
+
     @overload
     def iter(
         self,
@@ -711,9 +829,8 @@ class TemporalAgent(WrapperAgent[AgentDepsT, OutputDataT]):
         if workflow.in_workflow():
             if not self._temporal_overrides_active.get():
                 raise UserError(
-                    '`agent.iter()` cannot currently be used inside a Temporal workflow. '
-                    'Set an `event_stream_handler` on the agent and use `agent.run()` instead. '
-                    'Please file an issue if this is not sufficient for your use case.'
+                    '`agent.iter()` cannot be used inside a Temporal workflow. '
+                    'Set an `event_stream_handler` on the agent and use `agent.run()` instead.'
                 )
 
             if model is not None:

pydantic_ai/durable_exec/temporal/_model.py

@@ -128,6 +128,8 @@ class TemporalModel(WrapperModel):
         if not workflow.in_workflow():
             return await super().request(messages, model_settings, model_request_parameters)
 
+        self._validate_model_request_parameters(model_request_parameters)
+
         return await workflow.execute_activity(  # pyright: ignore[reportUnknownMemberType]
             activity=self.request_activity,
             arg=_RequestParams(
@@ -163,6 +165,8 @@ class TemporalModel(WrapperModel):
         # and that only calls `request_stream` if `event_stream_handler` is set.
         assert self.event_stream_handler is not None
 
+        self._validate_model_request_parameters(model_request_parameters)
+
         serialized_run_context = self.run_context_type.serialize_run_context(run_context)
         response = await workflow.execute_activity(  # pyright: ignore[reportUnknownMemberType]
             activity=self.request_stream_activity,
@@ -178,3 +182,7 @@ class TemporalModel(WrapperModel):
             **self.activity_config,
         )
         yield TemporalStreamedResponse(model_request_parameters, response)
+
+    def _validate_model_request_parameters(self, model_request_parameters: ModelRequestParameters) -> None:
+        if model_request_parameters.allow_image_output:
+            raise UserError('Image output is not supported with Temporal because of the 2MB payload size limit.')

pydantic_ai/format_prompt.py

@@ -3,6 +3,7 @@ from __future__ import annotations as _annotations
 from collections.abc import Iterable, Iterator, Mapping
 from dataclasses import asdict, dataclass, field, fields, is_dataclass
 from datetime import date
+from enum import Enum
 from typing import Any, Literal
 from xml.etree import ElementTree
 
@@ -26,8 +27,8 @@ def format_as_xml(
     This is useful since LLMs often find it easier to read semi-structured data (e.g. examples) as XML,
     rather than JSON etc.
 
-    Supports: `str`, `bytes`, `bytearray`, `bool`, `int`, `float`, `date`, `datetime`, `Mapping`,
-    `Iterable`, `dataclass`, and `BaseModel`.
+    Supports: `str`, `bytes`, `bytearray`, `bool`, `int`, `float`, `date`, `datetime`, `Enum`,
+    `Mapping`, `Iterable`, `dataclass`, and `BaseModel`.
 
     Args:
         obj: Python Object to serialize to XML.
@@ -101,7 +102,7 @@ class _ToXml:
             element.text = value
         elif isinstance(value, bytes | bytearray):
             element.text = value.decode(errors='ignore')
-        elif isinstance(value, bool | int | float):
+        elif isinstance(value, bool | int | float | Enum):
             element.text = str(value)
         elif isinstance(value, date):
             element.text = value.isoformat()

pydantic_ai/mcp.py

@@ -167,6 +167,10 @@ class MCPServer(AbstractToolset[Any], ABC):
     def id(self) -> str | None:
         return self._id
 
+    @id.setter
+    def id(self, value: str | None):
+        self._id = value
+
     @property
     def label(self) -> str:
         if self.id:
@@ -414,6 +418,9 @@ class MCPServer(AbstractToolset[Any], ABC):
         else:
             assert_never(resource)
 
+    def __eq__(self, value: object, /) -> bool:
+        return isinstance(value, MCPServer) and self.id == value.id and self.tool_prefix == value.tool_prefix
+
 
 class MCPServerStdio(MCPServer):
     """Runs an MCP server in a subprocess and communicates with it over stdin/stdout.
@@ -568,10 +575,10 @@ class MCPServerStdio(MCPServer):
         return f'{self.__class__.__name__}({", ".join(repr_args)})'
 
     def __eq__(self, value: object, /) -> bool:
-        if not isinstance(value, MCPServerStdio):
-            return False  # pragma: no cover
         return (
-            self.command == value.command
+            super().__eq__(value)
+            and isinstance(value, MCPServerStdio)
+            and self.command == value.command
             and self.args == value.args
             and self.env == value.env
             and self.cwd == value.cwd
@@ -809,9 +816,7 @@ class MCPServerSSE(_MCPServerHTTP):
         return sse_client  # pragma: no cover
 
     def __eq__(self, value: object, /) -> bool:
-        if not isinstance(value, MCPServerSSE):
-            return False  # pragma: no cover
-        return self.url == value.url
+        return super().__eq__(value) and isinstance(value, MCPServerSSE) and self.url == value.url
 
 
 @deprecated('The `MCPServerHTTP` class is deprecated, use `MCPServerSSE` instead.')
@@ -885,9 +890,7 @@ class MCPServerStreamableHTTP(_MCPServerHTTP):
         return streamablehttp_client  # pragma: no cover
 
     def __eq__(self, value: object, /) -> bool:
-        if not isinstance(value, MCPServerStreamableHTTP):
-            return False  # pragma: no cover
-        return self.url == value.url
+        return super().__eq__(value) and isinstance(value, MCPServerStreamableHTTP) and self.url == value.url
 
 
 ToolResult = (
@@ -964,4 +967,11 @@ def load_mcp_servers(config_path: str | Path) -> list[MCPServerStdio | MCPServer
         raise FileNotFoundError(f'Config file {config_path} not found')
 
     config = MCPServerConfig.model_validate_json(config_path.read_bytes())
-    return list(config.mcp_servers.values())
+
+    servers: list[MCPServerStdio | MCPServerStreamableHTTP | MCPServerSSE] = []
+    for name, server in config.mcp_servers.items():
+        server.id = name
+        server.tool_prefix = name
+        servers.append(server)
+
+    return servers

pydantic_ai/messages.py

@@ -13,7 +13,7 @@ import pydantic
 import pydantic_core
 from genai_prices import calc_price, types as genai_types
 from opentelemetry._events import Event  # pyright: ignore[reportPrivateImportUsage]
-from typing_extensions import deprecated
+from typing_extensions import Self, deprecated
 
 from . import _otel_messages, _utils
 from ._utils import generate_tool_call_id as _generate_tool_call_id, now_utc as _now_utc
@@ -461,7 +461,7 @@ class BinaryContent:
     """The media type of the binary data."""
 
     identifier: str
-    """Identifier for the binary content, such as a unique ID. generating one from the data if not explicitly set
+    """Identifier for the binary content, such as a unique ID.
     This identifier can be provided to the model in a message to allow it to refer to this file in a tool call argument,
     and the tool can look up the file in question by iterating over the message history and finding the matching `BinaryContent`.
 
@@ -496,6 +496,33 @@ class BinaryContent:
         self.vendor_metadata = vendor_metadata
         self.kind = kind
 
+    @staticmethod
+    def narrow_type(bc: BinaryContent) -> BinaryContent | BinaryImage:
+        """Narrow the type of the `BinaryContent` to `BinaryImage` if it's an image."""
+        if bc.is_image:
+            return BinaryImage(
+                data=bc.data,
+                media_type=bc.media_type,
+                identifier=bc.identifier,
+                vendor_metadata=bc.vendor_metadata,
+            )
+        else:
+            return bc  # pragma: no cover
+
+    @classmethod
+    def from_data_uri(cls, data_uri: str) -> Self:
+        """Create a `BinaryContent` from a data URI."""
+        prefix = 'data:'
+        if not data_uri.startswith(prefix):
+            raise ValueError('Data URI must start with "data:"')  # pragma: no cover
+        media_type, data = data_uri[len(prefix) :].split(';base64,', 1)
+        return cls(data=base64.b64decode(data), media_type=media_type)
+
+    @property
+    def data_uri(self) -> str:
+        """Convert the `BinaryContent` to a data URI."""
+        return f'data:{self.media_type};base64,{base64.b64encode(self.data).decode()}'
+
     @property
     def is_audio(self) -> bool:
         """Return `True` if the media type is an audio type."""
@@ -534,6 +561,24 @@ class BinaryContent:
     __repr__ = _utils.dataclasses_no_defaults_repr
 
 
+class BinaryImage(BinaryContent):
+    """Binary content that's guaranteed to be an image."""
+
+    def __init__(
+        self,
+        data: bytes,
+        *,
+        media_type: str,
+        identifier: str | None = None,
+        vendor_metadata: dict[str, Any] | None = None,
+        kind: Literal['binary'] = 'binary',
+    ):
+        super().__init__(data=data, media_type=media_type, identifier=identifier, vendor_metadata=vendor_metadata)
+
+        if not self.is_image:
+            raise ValueError('`BinaryImage` must be have a media type that starts with "image/"')  # pragma: no cover
+
+
 MultiModalContent = ImageUrl | AudioUrl | DocumentUrl | VideoUrl | BinaryContent
 UserContent: TypeAlias = str | MultiModalContent
 
@@ -934,6 +979,32 @@ class ThinkingPart:
     __repr__ = _utils.dataclasses_no_defaults_repr
 
 
+@dataclass(repr=False)
+class FilePart:
+    """A file response from a model."""
+
+    content: Annotated[BinaryContent, pydantic.AfterValidator(BinaryImage.narrow_type)]
+    """The file content of the response."""
+
+    _: KW_ONLY
+
+    id: str | None = None
+    """The identifier of the file part."""
+
+    provider_name: str | None = None
+    """The name of the provider that generated the response.
+    """
+
+    part_kind: Literal['file'] = 'file'
+    """Part type identifier, this is available on all parts as a discriminator."""
+
+    def has_content(self) -> bool:
+        """Return `True` if the file content is non-empty."""
+        return bool(self.content)  # pragma: no cover
+
+    __repr__ = _utils.dataclasses_no_defaults_repr
+
+
 @dataclass(repr=False)
 class BaseToolCallPart:
     """A tool call from a model."""
@@ -1016,7 +1087,7 @@ class BuiltinToolCallPart(BaseToolCallPart):
 
 
 ModelResponsePart = Annotated[
-    TextPart | ToolCallPart | BuiltinToolCallPart | BuiltinToolReturnPart | ThinkingPart,
+    TextPart | ToolCallPart | BuiltinToolCallPart | BuiltinToolReturnPart | ThinkingPart | FilePart,
     pydantic.Discriminator('part_kind'),
 ]
 """A message part returned by a model."""
@@ -1073,6 +1144,61 @@ class ModelResponse:
     finish_reason: FinishReason | None = None
     """Reason the model finished generating the response, normalized to OpenTelemetry values."""
 
+    @property
+    def text(self) -> str | None:
+        """Get the text in the response."""
+        texts: list[str] = []
+        last_part: ModelResponsePart | None = None
+        for part in self.parts:
+            if isinstance(part, TextPart):
+                # Adjacent text parts should be joined together, but if there are parts in between
+                # (like built-in tool calls) they should have newlines between them
+                if isinstance(last_part, TextPart):
+                    texts[-1] += part.content
+                else:
+                    texts.append(part.content)
+            last_part = part
+        if not texts:
+            return None
+
+        return '\n\n'.join(texts)
+
+    @property
+    def thinking(self) -> str | None:
+        """Get the thinking in the response."""
+        thinking_parts = [part.content for part in self.parts if isinstance(part, ThinkingPart)]
+        if not thinking_parts:
+            return None
+        return '\n\n'.join(thinking_parts)
+
+    @property
+    def files(self) -> list[BinaryContent]:
+        """Get the files in the response."""
+        return [part.content for part in self.parts if isinstance(part, FilePart)]
+
+    @property
+    def images(self) -> list[BinaryImage]:
+        """Get the images in the response."""
+        return [file for file in self.files if isinstance(file, BinaryImage)]
+
+    @property
+    def tool_calls(self) -> list[ToolCallPart]:
+        """Get the tool calls in the response."""
+        return [part for part in self.parts if isinstance(part, ToolCallPart)]
+
+    @property
+    def builtin_tool_calls(self) -> list[tuple[BuiltinToolCallPart, BuiltinToolReturnPart]]:
+        """Get the builtin tool calls and results in the response."""
+        calls = [part for part in self.parts if isinstance(part, BuiltinToolCallPart)]
+        if not calls:
+            return []
+        returns_by_id = {part.tool_call_id: part for part in self.parts if isinstance(part, BuiltinToolReturnPart)}
+        return [
+            (call_part, returns_by_id[call_part.tool_call_id])
+            for call_part in calls
+            if call_part.tool_call_id in returns_by_id
+        ]
+
     @deprecated('`price` is deprecated, use `cost` instead')
     def price(self) -> genai_types.PriceCalculation:  # pragma: no cover
         return self.cost()
@@ -1118,6 +1244,18 @@ class ModelResponse:
                 body.setdefault('content', []).append(
                     {'kind': kind, **({'text': part.content} if settings.include_content else {})}
                 )
+            elif isinstance(part, FilePart):
+                body.setdefault('content', []).append(
+                    {
+                        'kind': 'binary',
+                        'media_type': part.content.media_type,
+                        **(
+                            {'binary_content': base64.b64encode(part.content.data).decode()}
+                            if settings.include_content and settings.include_binary_content
+                            else {}
+                        ),
+                    }
+                )
 
         if content := body.get('content'):
             text_content = content[0].get('text')
@@ -1143,6 +1281,11 @@ class ModelResponse:
                         **({'content': part.content} if settings.include_content else {}),
                     )
                 )
+            elif isinstance(part, FilePart):
+                converted_part = _otel_messages.BinaryDataPart(type='binary', media_type=part.content.media_type)
+                if settings.include_content and settings.include_binary_content:
+                    converted_part['content'] = base64.b64encode(part.content.data).decode()
+                parts.append(converted_part)
             elif isinstance(part, BaseToolCallPart):
                 call_part = _otel_messages.ToolCallPart(type='tool_call', id=part.tool_call_id, name=part.tool_name)
                 if isinstance(part, BuiltinToolCallPart):
@@ -1511,6 +1654,9 @@ class FunctionToolResultEvent:
 
     _: KW_ONLY
 
+    content: str | Sequence[UserContent] | None = None
+    """The content that will be sent to the model as a UserPromptPart following the result."""
+
     event_kind: Literal['function_tool_result'] = 'function_tool_result'
     """Event type identifier, used as a discriminator."""
 

pydantic_ai/models/__init__.py

@@ -27,6 +27,8 @@ from .._run_context import RunContext
 from ..builtin_tools import AbstractBuiltinTool
 from ..exceptions import UserError
 from ..messages import (
+    BinaryImage,
+    FilePart,
     FileUrl,
     FinalResultEvent,
     FinishReason,
@@ -141,12 +143,20 @@ KnownModelName = TypeAliasType(
         'google-gla:gemini-2.0-flash',
         'google-gla:gemini-2.0-flash-lite',
         'google-gla:gemini-2.5-flash',
+        'google-gla:gemini-2.5-flash-preview-09-2025',
+        'google-gla:gemini-flash-latest',
         'google-gla:gemini-2.5-flash-lite',
+        'google-gla:gemini-2.5-flash-lite-preview-09-2025',
+        'google-gla:gemini-flash-lite-latest',
         'google-gla:gemini-2.5-pro',
         'google-vertex:gemini-2.0-flash',
         'google-vertex:gemini-2.0-flash-lite',
         'google-vertex:gemini-2.5-flash',
+        'google-vertex:gemini-2.5-flash-preview-09-2025',
+        'google-vertex:gemini-flash-latest',
         'google-vertex:gemini-2.5-flash-lite',
+        'google-vertex:gemini-2.5-flash-lite-preview-09-2025',
+        'google-vertex:gemini-flash-lite-latest',
         'google-vertex:gemini-2.5-pro',
         'grok:grok-4',
         'grok:grok-4-0709',
@@ -300,6 +310,7 @@ class ModelRequestParameters:
     output_object: OutputObjectDefinition | None = None
     output_tools: list[ToolDefinition] = field(default_factory=list)
     allow_text_output: bool = True
+    allow_image_output: bool = False
 
     @cached_property
     def tool_defs(self) -> dict[str, ToolDefinition]:
@@ -557,6 +568,7 @@ class StreamedResponse(ABC):
             finish_reason=self.finish_reason,
         )
 
+    # TODO (v2): Make this a property
     def usage(self) -> RequestUsage:
         """Get the usage of the response so far. This will not be the final usage until the stream is exhausted."""
         return self._usage
@@ -865,7 +877,9 @@ def _get_final_result_event(e: ModelResponseStreamEvent, params: ModelRequestPar
     """Return an appropriate FinalResultEvent if `e` corresponds to a part that will produce a final result."""
     if isinstance(e, PartStartEvent):
         new_part = e.part
-        if isinstance(new_part, TextPart) and params.allow_text_output:  # pragma: no branch
+        if (isinstance(new_part, TextPart) and params.allow_text_output) or (
+            isinstance(new_part, FilePart) and params.allow_image_output and isinstance(new_part.content, BinaryImage)
+        ):
             return FinalResultEvent(tool_name=None, tool_call_id=None)
         elif isinstance(new_part, ToolCallPart) and (tool_def := params.tool_defs.get(new_part.tool_name)):
             if tool_def.kind == 'output':

pydantic_ai/models/anthropic.py

@@ -20,6 +20,7 @@ from ..messages import (
     BuiltinToolCallPart,
     BuiltinToolReturnPart,
     DocumentUrl,
+    FilePart,
     FinishReason,
     ImageUrl,
     ModelMessage,
@@ -545,6 +546,9 @@ class AnthropicModel(Model):
                                         ),
                                     )
                                 )
+                    elif isinstance(response_part, FilePart):  # pragma: no cover
+                        # Files generated by models are not sent back to models that don't themselves generate files.
+                        pass
                     else:
                         assert_never(response_part)
                 if len(assistant_content_params) > 0:
@@ -693,17 +697,17 @@ class AnthropicStreamedResponse(StreamedResponse):
                     if maybe_event is not None:  # pragma: no branch
                         yield maybe_event
                 elif isinstance(current_block, BetaServerToolUseBlock):
-                    yield self._parts_manager.handle_builtin_tool_call_part(
+                    yield self._parts_manager.handle_part(
                         vendor_part_id=event.index,
                         part=_map_server_tool_use_block(current_block, self.provider_name),
                     )
                 elif isinstance(current_block, BetaWebSearchToolResultBlock):
-                    yield self._parts_manager.handle_builtin_tool_return_part(
+                    yield self._parts_manager.handle_part(
                         vendor_part_id=event.index,
                         part=_map_web_search_tool_result_block(current_block, self.provider_name),
                     )
                 elif isinstance(current_block, BetaCodeExecutionToolResultBlock):
-                    yield self._parts_manager.handle_builtin_tool_return_part(
+                    yield self._parts_manager.handle_part(
                         vendor_part_id=event.index,
                         part=_map_code_execution_tool_result_block(current_block, self.provider_name),
                     )

pydantic_ai/models/cohere.py

@@ -13,6 +13,7 @@ from .._utils import generate_tool_call_id as _generate_tool_call_id, guard_tool
 from ..messages import (
     BuiltinToolCallPart,
     BuiltinToolReturnPart,
+    FilePart,
     FinishReason,
     ModelMessage,
     ModelRequest,
@@ -255,6 +256,9 @@ class CohereModel(Model):
                     elif isinstance(item, BuiltinToolCallPart | BuiltinToolReturnPart):  # pragma: no cover
                         # This is currently never returned from cohere
                         pass
+                    elif isinstance(item, FilePart):  # pragma: no cover
+                        # Files generated by models are not sent back to models that don't themselves generate files.
+                        pass
                     else:
                         assert_never(item)
 

pydantic_ai/models/function.py

@@ -18,6 +18,7 @@ from ..messages import (
     BinaryContent,
     BuiltinToolCallPart,
     BuiltinToolReturnPart,
+    FilePart,
     ModelMessage,
     ModelRequest,
     ModelResponse,
@@ -319,12 +320,12 @@ class FunctionStreamedResponse(StreamedResponse):
                         if content := delta.args_as_json_str():  # pragma: no branch
                             response_tokens = _estimate_string_tokens(content)
                             self._usage += usage.RequestUsage(output_tokens=response_tokens)
-                        yield self._parts_manager.handle_builtin_tool_call_part(vendor_part_id=dtc_index, part=delta)
+                        yield self._parts_manager.handle_part(vendor_part_id=dtc_index, part=delta)
                     elif isinstance(delta, BuiltinToolReturnPart):
                         if content := delta.model_response_str():  # pragma: no branch
                             response_tokens = _estimate_string_tokens(content)
                             self._usage += usage.RequestUsage(output_tokens=response_tokens)
-                        yield self._parts_manager.handle_builtin_tool_return_part(vendor_part_id=dtc_index, part=delta)
+                        yield self._parts_manager.handle_part(vendor_part_id=dtc_index, part=delta)
                     else:
                         assert_never(delta)
 
@@ -371,10 +372,12 @@ def _estimate_usage(messages: Iterable[ModelMessage]) -> usage.RequestUsage:
                     response_tokens += _estimate_string_tokens(part.content)
                 elif isinstance(part, ToolCallPart):
                     response_tokens += 1 + _estimate_string_tokens(part.args_as_json_str())
-                elif isinstance(part, BuiltinToolCallPart):  # pragma: no cover
+                elif isinstance(part, BuiltinToolCallPart):
                     response_tokens += 1 + _estimate_string_tokens(part.args_as_json_str())
-                elif isinstance(part, BuiltinToolReturnPart):  # pragma: no cover
+                elif isinstance(part, BuiltinToolReturnPart):
                     response_tokens += _estimate_string_tokens(part.model_response_str())
+                elif isinstance(part, FilePart):
+                    response_tokens += _estimate_string_tokens([part.content])
                 else:
                     assert_never(part)
         else: