> ## Documentation Index > Fetch the complete documentation index at: https://docs.platform.aiplanet.com/llms.txt > Use this file to discover all available pages before exploring further. # Running Workflows > How runs, threads, and inputs work when you execute a workflow. Once a workflow is built, you run it to get results. This page explains what happens during a run and the concepts involved. ## Runs A **run** is a single execution of a workflow. Every time a workflow is triggered — from the Playground, by your team, or through the API — a new run is created. Each run has: * A **status** — pending, running, completed, failed, or cancelled. * The **inputs** it was given. * The **outputs** it produced. * A full **trace** of every node that executed. See [Tracing](/running-monitoring/tracing). ## Inputs A workflow's **Input** node defines what information a run needs. When you start a run, you provide a value for each input parameter. Required parameters must be filled in; the run won't start without them. ## Threads A **thread** is a conversation session that groups related runs together. Runs in the same thread share context, which lets an agent with [memory](/knowledge-memory/memory-configurations) refer back to earlier messages. * Use a **new thread** for an independent task with no prior context. * Use an **existing thread** to continue a conversation. ## Ways to run a workflow Interactive testing against the workflow's active deployment. Triggered from your own applications. Requires a deployed workflow and an API key. ## Streaming results Runs stream their output as it's produced, so you see results progressively rather than waiting for the whole run to finish. In the Playground this appears as the response building up live; through the API it's available as a stream of events. ## Cancelling a run A run that's in progress can be cancelled if you no longer need its result. ## Next steps Inspect exactly what happened in any run. Understand concurrency and usage limits.