The actions taken by the agent.
Optionalauthor"user" or the name of the agent, indicating who appended the event to the session.
OptionalbranchThe branch of the event. The format is like agent_1.agent_2.agent_3, where agent_1 is the parent of agent_2, and agent_2 is the parent of agent_3.
Branch is used when multiple sub-agent shouldn't see their peer agents' conversation history.
OptionalcitationThe citation metadata of the response.
The summarized content of the compacted events.
OptionalcontentThe content of the response.
OptionalcustomThe custom metadata of the LlmResponse. An optional key-value pair to label an LlmResponse. NOTE: the entire object must be JSON serializable.
The end time of the context that was compacted.
OptionalerrorError code if the response is an error. Code varies by model.
OptionalerrorError message if the response is an error.
OptionalfinishThe finish reason of the response.
OptionalgoServer-side signal that the live connection will be closed soon. The caller should reconnect using the latest session resumption handle.
OptionalgroundingThe grounding metadata of the response.
The unique identifier of the event. Do not assign the ID. It will be assigned by the session.
OptionalinputAudio transcription of user input.
OptionalinteractionThe interaction ID returned by the model, if any.
OptionalinterruptedFlag indicating that LLM was interrupted when generating the content. Usually it's due to user interruption during a bidi streaming.
The invocation ID of the event. Should be non-empty before appending to a session.
ReadonlyisIdentifies this event as a compacted event.
OptionalisIdentifies this compacted event as the persistent context scratchpad.
OptionalliveThe session ID of the Live session.
OptionalliveThe session resumption update of the LlmResponse
OptionallongSet of ids of the long running function calls. Agent client will know from this field about which function call is long running. Only valid for function call event
OptionalmodelThe model version used to generate the response.
OptionaloutputAudio transcription of model output.
OptionalpartialIndicates whether the text content is part of a unfinished text stream. Only used for streaming mode and when the content is plain text.
The start time of the context that was compacted.
The timestamp of the event.
OptionalturnIndicates whether the response from the model is complete. Only used for streaming mode.
OptionalusageThe usage metadata of the LlmResponse.
A specialized Event type that represents a synthesized summary of past events. This is used to compress session history without losing critical context.