google.generativeai.ChatSession

View source on GitHub

Contains an ongoing conversation with the model.

google.generativeai.ChatSession(
    model: GenerativeModel,
    history: (Iterable[content_types.StrictContentType] | None) = None,
    enable_automatic_function_calling: bool = False
)

>>> model = genai.GenerativeModel('models/gemini-1.5-flash')
>>> chat = model.start_chat()
>>> response = chat.send_message("Hello")
>>> print(response.text)
>>> response = chat.send_message("Hello again")
>>> print(response.text)
>>> response = chat.send_message(...

This ChatSession object collects the messages sent and received, in its ChatSession.history attribute.

Arguments
model	The model to use in the chat.
history	A chat history to initialize the object with.

Attributes
history	The chat history.
last	returns the last received genai.GenerateContentResponse

Methods

rewind

View source

rewind() -> tuple[protos.Content, protos.Content]

Removes the last request/response pair from the chat history.

send_message

View source

send_message(
    content: content_types.ContentType,
    *,
    generation_config: generation_types.GenerationConfigType = None,
    safety_settings: safety_types.SafetySettingOptions = None,
    stream: bool = False,
    tools: (content_types.FunctionLibraryType | None) = None,
    tool_config: (content_types.ToolConfigType | None) = None,
    request_options: (helper_types.RequestOptionsType | None) = None
) -> generation_types.GenerateContentResponse

Sends the conversation history with the added message and returns the model’s response.

Appends the request and response to the conversation history.

>>> model = genai.GenerativeModel('models/gemini-1.5-flash')
>>> chat = model.start_chat()
>>> response = chat.send_message("Hello")
>>> print(response.text)
"Hello! How can I assist you today?"
>>> len(chat.history)
2

Call it with stream=True to receive response chunks as they are generated:

>>> chat = model.start_chat()
>>> response = chat.send_message("Explain quantum physics", stream=True)
>>> for chunk in response:
...   print(chunk.text, end='')

Once iteration over chunks is complete, the response and ChatSession are in states identical to the stream=False case. Some properties are not available until iteration is complete.

Like GenerativeModel.generate_content this method lets you override the model’s generation_config and safety_settings.

Arguments
content	The message contents.
generation_config	Overrides for the model’s generation config.
safety_settings	Overrides for the model’s safety settings.
stream	If True, yield response chunks as they are generated.

send_message_async

View source

send_message_async(
    content,
    *,
    generation_config=None,
    safety_settings=None,
    stream=False,
    tools=None,
    tool_config=None,
    request_options=None
)

The async version of ChatSession.send_message.