Runs domain

class yandexcloudmlsdk.runs.domain.AsyncRunsclass yandex_cloud_ml_sdk._runs.domain.AsyncRuns

Class for Runs operations. Provides core functionality for managing assistant execution in streams.

For usage examples see runs example.

async get(run_id, *, timeout=60)

Get a run by ID.

Parameters	run_id (str) – Run ID timeout (float) – The timeout, or the maximum time to wait for the request to complete in seconds. Defaults to 60 seconds.
Return type	AsyncRun

async get_last_by_thread(thread, *, timeout=60)

Get the last run for a thread.

Parameters	thread (str | BaseThread) – Thread ID or instance timeout (float) – The timeout, or the maximum time to wait for the request to complete in seconds. Defaults to 60 seconds.
Return type	AsyncRun

async list(*, page_size=Undefined, timeout=60)

List runs. Returns an async iterator to retrieve all runs.

Parameters	page_size (int | Undefined) – Number of items per page. Larger values reduce the number of network calls but increase memory consumption per request. timeout (float) – The timeout, or the maximum time to wait for each network request to complete in seconds. Defaults to 60 seconds. This affects the network behavior but not the total time for iteration.
Return type	AsyncIterator[AsyncRun]

class yandexcloudmlsdk.runs.run.AsyncRunclass yandex_cloud_ml_sdk._runs.run.AsyncRun

Asynchronous implementation of Run operations.

Represents a server-side Run object that has been started by an assistant on a specific thread. It implements the Operation interface, allowing you to monitor the Run’s execution on the server, track its progress, and retrieve its results.

The AsyncRun provides asynchronous methods to: - Listen to real-time events from the running assistant - Submit tool execution results back to continue the conversation - Monitor the Run’s status and retrieve final results - Handle the complete lifecycle of an assistant conversation session

async listen(*, events_start_idx=0, timeout=60)

Listen to run events stream.

Parameters	events_start_idx (int) – Starting event index timeout (float) – The timeout, or the maximum time to wait for the request to complete in seconds. Defaults to 60 seconds.
Return type	AsyncIterator[RunStreamEvent[AsyncToolCall]]

async submit_tool_results(tool_results, *, timeout=60)

Submit tool execution results to continue the run.

Parameters	tool_results (FunctionResultDict | Iterable[FunctionResultDict]) – Tool call results to submit timeout (float) – The timeout, or the maximum time to wait for the request to complete in seconds. Defaults to 60 seconds.
Return type	None

async cancel(*, timeout=60)

Parameters	timeout (float)
Return type	None

property custom_max_prompt_tokens: int | None

Get max prompt tokens from truncation options if set.

async get_result(*, timeout=60)

Parameters	timeout (float)
Return type	AnyResultTypeT_co

async get_status(*, timeout=60)

Parameters	timeout (float)
Return type	OperationStatusTypeT

async wait(*, timeout=60, poll_timeout=None, poll_interval=None)

Parameters	timeout (float) poll_timeout (int | None) poll_interval (float | None)
Return type	AnyResultTypeT_co

id: str

Unique run identifier

assistant_id: str

ID of the assistant used

thread_id: str

ID of the thread used

created_by: str

Creator of the run

created_at: datetime

Creation timestamp

labels: dict[str, str] | None

Optional metadata labels

custom_temperature: float | None

Custom temperature setting

custom_max_tokens: int | None

Custom max tokens setting

custom_prompt_truncation_options: PromptTruncationOptions | None

Custom prompt truncation options

custom_response_format: ResponseType | None

Custom response format