Stateless Create Run

Create a stateless background run.

Generates an ephemeral thread, delegates to the threaded create_run endpoint, and schedules cleanup as a background task (unless on_completion="keep").

POST /runs

Stateless Create Run

cURL

curl --request POST \
  --url https://api.example.com/runs \
  --header 'Content-Type: application/json' \
  --data '
{
  "assistant_id": "<string>",
  "input": {},
  "config": {},
  "context": {},
  "checkpoint": {},
  "stream": false,
  "stream_mode": "<string>",
  "on_disconnect": "<string>",
  "on_completion": "delete",
  "multitask_strategy": "<string>",
  "command": {},
  "interrupt_before": "<string>",
  "interrupt_after": "<string>",
  "stream_subgraphs": false,
  "metadata": {}
}
'

200

{
  "run_id": "<string>",
  "thread_id": "<string>",
  "assistant_id": "<string>",
  "input": {},
  "user_id": "<string>",
  "created_at": "2023-11-07T05:31:56Z",
  "updated_at": "2023-11-07T05:31:56Z",
  "status": "pending",
  "output": {},
  "error_message": "<string>",
  "config": {},
  "context": {}
}

Body

application/json

Request model for creating runs

assistant_id · string · required

Assistant to execute

input · Input · object

Input data for the run. Optional when resuming from a checkpoint.

config · Config · object

Execution config

context · Context · object

Execution context

checkpoint · Checkpoint · object

Checkpoint configuration (e.g., {'checkpoint_id': '...', 'checkpoint_ns': ''})

stream · boolean · default: false

Enable streaming response

stream_mode · string string[]

Requested stream mode(s)

on_disconnect · string | null

Behavior on client disconnect: 'cancel' (default) or 'continue'.

on_completion · enum<string> | null

Behavior after stateless run completes: 'delete' (default) removes the ephemeral thread, 'keep' preserves it.

Available options: delete, keep

multitask_strategy · string | null

Strategy for handling concurrent runs on same thread: 'reject', 'interrupt', 'rollback', or 'enqueue'.

command · Command · object

Command for resuming interrupted runs with state updates or navigation

interrupt_before · string string[]

Nodes to interrupt immediately before they get executed. Use '*' for all nodes.

interrupt_after · string string[]

Nodes to interrupt immediately after they get executed. Use '*' for all nodes.

stream_subgraphs · boolean | null · default: false

Whether to include subgraph events in streaming. When True, includes events from all subgraphs. When False (default when None), excludes subgraph events. Defaults to False for backwards compatibility.

metadata · Metadata · object

Request metadata (e.g., from_studio flag)

Response

200

application/json

Successful Response

Run entity model

Status values: pending, running, error, success, timeout, interrupted

run_id · string · required

Unique identifier for the run.

thread_id · string · required

Thread this run belongs to.

assistant_id · string · required

Assistant that is executing this run.

input · Input · object · required

Input data provided to the run.

user_id · string · required

Identifier of the user who owns this run.

created_at · string<date-time> · required

Timestamp when the run was created.

updated_at · string<date-time> · required

Timestamp when the run was last updated.

status · string · default: pending

Current run status: pending, running, error, success, timeout, or interrupted.

output · Output · object

Final output produced by the run, or null if not yet complete.

error_message · string | null

Error message if the run failed.

config · Config · object

Configuration passed to the graph at runtime.

context · Context · object

Context variables available during execution.