Skip to main content
What this builds. The async counterpart to a regular pipeline.run(...) — fetch by name, then call await pipeline.arun(input_data) so it composes with other awaitable work. You’ll end up with. The same result dict you’d get from pipeline.run(...), printed once the awaited call completes. 
import asyncio
import vectorshift

from vectorshift.pipeline import Pipeline, InputNode, OutputNode, LlmNode

vectorshift.api_key = "your api key here"


pipeline = Pipeline.fetch(name="your pipeline name here")

input_data = {"input_0": "Hello, how are you?"}

result = asyncio.run(pipeline.arun(input_data))
print(result)

​
Expected output

{'outputs': {'output_0': '...'}, 'status': 'success', 'run_id': '...'}
arun(...) returns the same shape as run(...); only the execution path differs.

​
See also

Background runs (async)

Start a run in the background instead of awaiting it.

Streaming (async)

async for chunk in pipeline.astream(...).

Pipeline reference

Full method surface.