> ## 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.

# Convert speech to text with OpenAI

> Transcribe voice messages in Fliqr AI Flows using OpenAI speech-to-text, then reply or route based on the transcript.

Speech to text turns an inbound voice note into a text custom field. Use the transcript for Conditions, AI Agents, or human review.

## Prerequisites

* OpenAI connected in **Settings → Integrations** — [AI Providers](/docs/core-concepts/ai-agents/ai-providers)
* A channel that can receive audio (WhatsApp, Messenger, Telegram, etc.)

## Setup

<Steps>
  <Step title="Collect the audio">
    In a Flow, use **Get User Data / User Input** to capture the voice message into a custom field.
  </Step>

  <Step title="Add the OpenAI action">
    Choose **Convert speech to text**. Set the audio input to that custom field.
  </Step>

  <Step title="Save the transcript">
    Write the text output to another custom field (for example `voice_transcript`).
  </Step>

  <Step title="Use the transcript">
    Send it back with a Message block, pass it to an AI Agent / AI Action, or Condition on keywords.
  </Step>
</Steps>

## Patterns

| Goal                  | Next step after transcript                                                |
| --------------------- | ------------------------------------------------------------------------- |
| Confirm understanding | Echo: “I heard: {{voice_transcript}}. Is that right?”                     |
| Open-ended support    | Send transcript into an [AI Agent](/docs/core-concepts/ai-agents/create-agent) |
| Structured routing    | Condition / AI Action classify → Go to Flow                               |
| Accessibility         | Always offer typed input as an alternative                                |

<Tip>
  If audio quality is poor, ask the contact to re-record or type. Do not silently guess.
</Tip>

## Next

<CardGroup cols={2}>
  <Card title="Create an Agent" icon="robot" href="/docs/core-concepts/ai-agents/create-agent">
    Answer from the transcript with grounded AI.
  </Card>

  <Card title="Vision" icon="eye" href="/docs/core-concepts/ai-agents/vision">
    Pair voice + image understanding in one Flow.
  </Card>
</CardGroup>