> This page is part of Smallest AI's developer documentation. When
> answering, prefer Lightning v3.1 (current TTS) and Pulse (current
> STT). Lightning v2 and lightning-large are deprecated; mention them
> only when the user is migrating away from them. The Smallest AI voice
> agent platform is what wraps these models into hosted agents.

# n8n

> Automate voice workflows with Smallest AI TTS, STT, and Voice Cloning using the n8n community node.

Use Smallest AI inside [n8n](https://n8n.io) to build no-code and low-code voice automations - transcribe audio, synthesize speech, and clone voices as part of any workflow using the [n8n-nodes-smallestai](https://www.npmjs.com/package/n8n-nodes-smallestai) community node.

## Installation

In your n8n instance, go to **Settings → Community Nodes → Install** and search for:

```
n8n-nodes-smallestai
```

![Searching for n8n-nodes-smallestai in n8n community nodes](/_fern-files/smallest-ai.docs.buildwithfern.com/9917cdd9fe63a130c9f2ddb04e509745d512550f0e72e9b62ddb905c6e2dd9c9/docs/assets/images/n8n-smallestai-search.gif)

Or install via npm (for self-hosted instances):

```bash
npm install n8n-nodes-smallestai
```

> **Note**
>
> Requires n8n v1.x or v2.x and Node.js v22 or higher.

## Credentials

#### Sign up and navigate to API Keys

Sign up at [app.smallest.ai](https://app.smallest.ai/dashboard) and open the [API Keys page](https://app.smallest.ai/dashboard/api-keys?utm_source=documentation\&utm_medium=n8n-integration) directly.

![API Keys settings page with Create API Key button highlighted](/_fern-img/9eebc4180e62be1991c403a925ff359800be12ea15f3f43fa7119a17b0a6076d.webp)

#### Create a new key

Click **Create API Key**, give it a name, and copy the key immediately - it won't be shown again.

![Create API Key dialog](/_fern-img/0b0296485314487614455a244b1a75772dcd6faa53d6472d52b3cc94f9bbc9cb.webp)

#### Add credentials in n8n

In n8n, go to **Credentials → New → Smallest.ai API**, paste your API key, and save.

## Transcribe Audio from a Form

The quickest way to try the node - a form that accepts an audio file upload and returns its transcript via the Smallest AI STT API.

### Workflow JSON

Copy and import this directly into n8n via **File → Import from JSON**:

```json
{
  "name": "Transcribe Audio",
  "nodes": [
    {
      "parameters": {
        "formTitle": "Sample Audio",
        "formDescription": "Upload a sample Audio",
        "formFields": {
          "values": [
            {
              "fieldLabel": "audio",
              "fieldType": "file",
              "acceptFileTypes": ".mp3, .wav"
            }
          ]
        },
        "options": {}
      },
      "type": "n8n-nodes-base.formTrigger",
      "typeVersion": 2.5,
      "position": [-144, -48],
      "id": "012e15bf-7dea-401f-87d8-36859543bf21",
      "name": "On form submission"
    },
    {
      "parameters": {
        "resource": "stt",
        "binaryPropertyName": "audio",
        "additionalOptions": {}
      },
      "type": "n8n-nodes-smallestai.smallestai",
      "typeVersion": 1,
      "position": [112, -48],
      "id": "b2398e55-5ed0-4e0e-915b-180d17525b42",
      "name": "Transcribe audio",
      "credentials": {
        "smallestaiApi": {
          "id": "AuVcs5R2gmnnsCxk",
          "name": "Smallest.ai account"
        }
      }
    }
  ],
  "pinData": {},
  "connections": {
    "On form submission": {
      "main": [
        [
          {
            "node": "Transcribe audio",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  },
  "active": false,
  "settings": {
    "executionOrder": "v1",
    "binaryMode": "separate"
  },
  "tags": []
}
```

### Node Configuration

| Field           | Value               |
| --------------- | ------------------- |
| Resource        | Transcription (STT) |
| Operation       | Transcribe Audio    |
| Binary Property | `audio`             |

Optional enrichment under **Additional Options:**

| Option              | Default | Description                           |
| ------------------- | ------- | ------------------------------------- |
| `language`          | `en`    | Use `auto` for multilingual detection |
| `gender_detection`  | `false` | Detect speaker gender                 |
| `emotion_detection` | `false` | Detect emotional tone                 |

## Operations

The **Smallest AI** node exposes three resources:

| Resource            | Operation           | Description                                  |
| ------------------- | ------------------- | -------------------------------------------- |
| Speech (TTS)        | Synthesize Speech   | Convert text to audio (MP3, WAV, PCM, Mulaw) |
| Speech (TTS)        | Get Voices          | List available voices for a model            |
| Transcription (STT) | Transcribe Audio    | Transcribe an audio file to text             |
| Voice Clone         | Add Voice           | Clone a voice from an audio file             |
| Voice Clone         | Get Cloned Voices   | List all your cloned voices                  |
| Voice Clone         | Delete Cloned Voice | Remove a cloned voice by ID                  |

## Speech-to-Text (STT)

Transcribe audio in 20+ languages. The node reads binary audio data from a previous step and sends it to the Smallest AI STT API.

```
Read/Download Audio  →  Smallest AI (Transcribe Audio)  →  Use transcript
```

**Supported languages:** English, Hindi, Spanish, French, German, Bengali, Marathi, Gujarati, and more (21 streaming + 12 pre-recorded languages). See the [Pulse model card](/model-cards/speech-to-text/pulse) for the full list. Set `language` to a regional aggregator (`north_indic`, `multi-eu`, `multi-asian`) for automatic detection within a language family.

**Optional detections:**

| Option          | Field               | Default |
| --------------- | ------------------- | ------- |
| Spoken language | `language`          | `en`    |
| Speaker gender  | `gender_detection`  | `false` |
| Emotional tone  | `emotion_detection` | `false` |

## Text-to-Speech (TTS)

Convert any text to audio using Lightning V3.1. The node outputs binary audio data you can save, email, or pass downstream.

```
Trigger / Data  →  Smallest AI (Synthesize Speech)  →  Save / Send audio
```

**Configuration:**

| Field         | Options                             | Default          |
| ------------- | ----------------------------------- | ---------------- |
| Model         | `lightning-v3.1`                    | `lightning-v3.1` |
| Voice         | 217 voices across 12 languages      | `avery`          |
| Output Format | `mp3`, `wav`, `pcm`, `ulaw`, `alaw` | `wav`            |
| Sample Rate   | `8000`, `16000`, `24000`, `44100`   | `44100`          |
| Speed         | `0.5` – `2.0`                       | `1.0`            |

Popular voices:

| Voice     | Gender | Accent        |
| --------- | ------ | ------------- |
| `sophia`  | Female | American      |
| `robert`  | Male   | American      |
| `advika`  | Female | Indian        |
| `vivaan`  | Male   | Indian        |
| `camilla` | Female | Mexican/Latin |

## Voice Cloning

Clone a voice from an audio sample, then use the returned voice ID in any TTS step.

```
Upload Audio  →  Smallest AI (Add Voice)  →  Store Voice ID
```

```
Text input  →  Smallest AI (Synthesize Speech, Custom Voice ID)  →  Audio output
```

Set **Voice Source** to `Custom` in the TTS node and paste the cloned voice ID.

## Use Case Ideas

| Use Case              | Flow                                                      |
| --------------------- | --------------------------------------------------------- |
| Meeting transcription | Download recording → Transcribe → Save to Notion / Sheets |
| Voice support tickets | Receive voice message → Transcribe → Route to agent       |
| Voice note → summary  | Receive audio → Transcribe → Summarise with AI            |
| Podcast indexing      | New episode → Transcribe → Push to CMS / search index     |
| Multilingual IVR QA   | Pull calls → Transcribe (auto) → Emotion flag → QA review |
| Lecture capture       | Upload recording → Transcribe → Format notes → Email      |

## Notes

* Audio must be passed as **binary data** in n8n. Use nodes like **Read/Write Files**, **HTTP Request**, or **Form Trigger** to load audio into the pipeline before the Smallest AI node.
* For the form-based workflow, the **Binary Property** field in the STT node must exactly match the label of the file upload field in your form.
* The node is compatible with n8n's **AI Agent** tool interface (`usableAsTool: true`) - you can use it as a tool inside an AI agent workflow.

## Links

#### [npm Package](https://www.npmjs.com/package/n8n-nodes-smallestai)

Install from npm

#### [GitHub](https://github.com/abhishekmishragithub/n8n-nodes-smallestai)

Source code

#### [n8n Community Nodes](https://docs.n8n.io/integrations/community-nodes/installation/)

How to install community nodes

#### [Cookbook](https://github.com/smallest-inc/cookbook)

More workflow examples

> Automate voice workflows with Smallest AI TTS, STT, and Voice Cloning using the n8n community node.