# Text to Audio

> Generate a sound effect or an ambience from a text description with ElevenLabs Sound Effects, up to 22 seconds long, with gapless loops for backgrounds.

Source: https://nodaro.ai/docs/nodes/audio/text-to-audio

The **Text to Audio** node generates a sound effect or an ambience from a text description. It runs the ElevenLabs Sound Effects model: you describe the sound, set its length, and the node returns an audio clip. Turn on **Loop** for a background, such as rain or a crowd, that repeats without a seam.

- Found in: Audio › Sound Effects
- Output: audio
- API type: `text-to-audio`

## When to use it
- Sound effects for a video: footsteps, a door, an impact.
- Ambient backgrounds: a forest, a city street, the ocean.
- Interface and notification sounds.
- Looping beds that repeat without a gap, under a podcast or a video.
- Sound textures for motion graphics.

To add matching sound to an existing video in one step, use [Video SFX](https://nodaro.ai/docs/nodes/video/video-sfx).

## Quick start
### Add the node

Press Tab on the canvas and choose **Audio › Sound Effects › Text to Audio**.

### Describe the sound

Open the settings panel and write the **Prompt**, for example "heavy rain on a tin roof". Describe the sound itself, not the story around it.

### Set the length

Set **Duration (seconds)**, from 0.5 to 22. For a background that repeats, set **Loop** to **On (seamless loop)**.

### Run it

Click **Run** on the node. The sound appears on the node with a player.

Workflow: Two generated sounds are mixed into one soundscape and merged onto a video.

- Text to Audio → Mix Audio (audio)
- Text to Audio → Mix Audio (audio)
- Mix Audio → Merge Video & Audio (audio)
- Generate Video → Merge Video & Audio (video)

## Inputs
| Input | Accepts | What it does |
| --- | --- | --- |
| **Prompt** | Text nodes, such as Text and Prompt | The description of the sound, instead of the Prompt field. |
| **Audio style** | Audio pickers, such as Music Mood and Tone | Adds the pickers' wording to the description. |

The output, **Audio**, is the URL of the generated sound.

## Settings
| Setting | What it does |
| --- | --- |
| **Prompt** | The description of the sound, up to 450 characters. |
| **Provider** | The sound model. **ElevenLabs SFX v2** is the only choice. |
| **Duration (seconds)** | From 0.5 to 22 seconds, in steps of 0.5. The default is 10. |
| **Loop** | **Off** (the default) or **On (seamless loop)**. On makes a clip whose end joins its start without a gap. |
| **Prompt Influence** | From 0 (more random) to 1 (more faithful). The default is 0.3. How literally the model follows the description. |
| **Pre & post text** | Text that is always added before and after the description. It is hidden from people who use your workflow as an app. See [Prompt pre and post text](https://nodaro.ai/docs/concepts/prompt-pre-post-text). |

## Models
| Model | Maker | Modes | Credits | Details |
| --- | --- | --- | --- | --- |
| [ElevenLabs Sound Effects](https://nodaro.ai/docs/models/audio/elevenlabs-sound-effects) | ElevenLabs | Sound effects | 3 | Generate short sound effects from a text prompt. |

The price is per generation. **Duration** does not change it.

## Tips
- **Describe the sound, not the scene.** "Heavy rain on a tin roof" works better than "it was a stormy night".
- **Keep Prompt Influence low.** Values from 0.2 to 0.4 sound the most natural. Raise it only when the result drifts from your description.
- **Loop continuous sounds.** Loop works best for steady ambience, such as wind, rain or a crowd, and less well for single hits.
- **Layer, then mix.** Generate each part of a soundscape in its own node, then combine them with [Mix Audio](https://nodaro.ai/docs/nodes/audio/mix-audio) and set each level.

## From the API
The MCP tool is `text_to_audio`. See the [MCP tools](https://nodaro.ai/docs/mcp/tools).

## Frequently asked questions

### How long can a generated sound effect be?

From 0.5 to 22 seconds, in steps of 0.5 seconds. The default is 10 seconds. For a longer background, turn on Loop and repeat the clip.

### How long can the description be?

Up to 450 characters. Describe one sound at a time, and build a complex soundscape from several nodes mixed together.

### Does a longer sound cost more?

No. The price is per generation, and the Duration setting does not change it.

### Can Text to Audio make speech or music?

It is made for sound effects and ambience, not speech. For speech, use Text to Speech. For music, use Suno Create Music or Generate Music.

### What does Prompt Influence do?

It sets how literally the model follows your description, from 0 to 1. The default is 0.3. Lower values allow a freer interpretation; raise the value only when the result drifts too far from the description.