> ## Documentation Index
> Fetch the complete documentation index at: https://fliqr.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Use AI Actions inside a Flow

> Call the LLM for a single step inside a Fliqr AI Flow — classify intent, summarize input, or extract fields without a full AI Agent.

An **AI Action** runs one model call inside an otherwise deterministic Flow. Use it when you need language understanding at a single step, then return to scripted branches.

## AI Action vs AI Agent

|          | AI Action                                  | [AI Agent](/docs/core-concepts/ai-agents) |
| -------- | ------------------------------------------ | ------------------------------------ |
| Scope    | One step in a Flow                         | Multi-turn open conversation         |
| Control  | You decide the next block after the result | Model drives the dialogue            |
| Best for | Classify, extract, summarize, score        | Open Q\&A, troubleshooting           |

## Common uses

* Classify free text (“billing” vs “shipping”) → [Condition](/docs/core-concepts/flows/conditions)
* Extract structured values into custom fields (order ID, city)
* Summarize a long user message before human handoff
* Score lead quality and tag the contact

## Build the step

<Steps>
  <Step title="Capture input">
    Use [User Input](/docs/core-concepts/flows/collect-data) or the latest inbound message as the model input.
  </Step>

  <Step title="Add AI Action">
    Configure the prompt, model/provider settings available in your workspace, and where to store the result (custom field).
  </Step>

  <Step title="Branch on the result">
    Add a Condition (or Go to Flow) that reads the saved field.
  </Step>

  <Step title="Add a fallback">
    If the model returns an unexpected value, send a clarification message or hand off to a human / full Agent.
  </Step>
</Steps>

<Tip>
  Keep prompts narrow and output formats strict (for example `Reply with exactly one of: billing, shipping, other`). Tight outputs make Conditions reliable.
</Tip>

## Cost and latency

AI Actions consume model tokens and add latency versus pure scripted blocks. Use them where language flexibility matters; prefer Quick Replies for closed choices.

## Next

<CardGroup cols={2}>
  <Card title="AI Agents" icon="robot" href="/docs/core-concepts/ai-agents">
    Full conversational agents when you need multi-turn reasoning.
  </Card>

  <Card title="Recipes" icon="book" href="/docs/core-concepts/flows/recipes">
    Example Flow patterns that mix script and AI.
  </Card>
</CardGroup>