> For the complete documentation index, see [llms.txt](https://docs.agenticflow.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.agenticflow.ai/reference/nodes/text_to_speech.md).
# Text to Speech
**Action ID:** `text_to_speech`
## Description
Convert text to natural-sounding speech audio using advanced text-to-speech technology. This node supports multiple voice providers including Azure, OpenAI, Google, and AWS Polly, offering a wide selection of voices in different languages and styles.
## Provider
**PixelML**
## Connection
| Name | Description | Required | Category |
| ------------------ | ------------------------------------------- | :------: | -------- |
| PixelML Connection | The PixelML connection to call PixelML API. | ✓ | pixelml |
## Input Parameters
| Name | Type | Required | Default | Description |
| ----- | ------------- | :------: | ------- | ----------------------------------------------------------------------------------------------------------------------------- |
| voice | string (enum) | ✓ | - | Voice to use for text to speech. Choose from a variety of voices across multiple providers (Azure, OpenAI, Google, AWS Polly) |
| text | string | ✓ | - | Text to convert to speech |
View JSON Schema
```json
{
"$defs": {
"TextToSpeechVoices": {
"description": "Text to speech voices.",
"enum": [
"Hoai_My_Neural",
"Nam_Minh_Neural",
"Ava_Neural",
"Andrew_Neural",
"Emma_Neural",
"Brian_Neural",
"Alloy",
"Echo",
"Fable",
"Onyx",
"Nova",
"Shimmer",
"en_US_Casual_K",
"en_US_Journey_D",
"en_US_Journey_F",
"en_US_Journey_O",
"en_US_Neural2_A",
"en_US_Neural2_C",
"en_US_Neural2_D",
"en_US_Neural2_E",
"en_US_Neural2_F",
"en_US_Neural2_G",
"en_US_Neural2_H",
"en_US_Neural2_I",
"en_US_Neural2_J",
"en_US_News_K",
"en_US_News_L",
"en_US_News_N",
"en_US_Polyglot_1",
"en_US_Standard_A",
"en_US_Standard_B",
"en_US_Standard_C",
"en_US_Standard_D",
"en_US_Standard_E",
"en_US_Standard_F",
"en_US_Standard_G",
"en_US_Standard_H",
"en_US_Standard_I",
"en_US_Standard_J",
"en_US_Studio_O",
"en_US_Studio_Q",
"en_US_Wavenet_A",
"en_US_Wavenet_B",
"en_US_Wavenet_C",
"en_US_Wavenet_D",
"en_US_Wavenet_E",
"en_US_Wavenet_F",
"en_US_Wavenet-G",
"en_US_Wavenet-H",
"en_US_Wavenet-I",
"en_US_Wavenet-J",
"vi_VN_Neural2_A",
"vi_VN_Neural2_D",
"vi_VN_Standard_A",
"vi_VN_Standard_B",
"vi_VN_Standard_C",
"vi_VN_Standard_D",
"vi_VN_Wavenet_A",
"vi_VN_Wavenet_B",
"vi_VN_Wavenet_C",
"vi_VN_Wavenet_D",
"Danielle",
"Gregory",
"Ivy",
"Joanna",
"Kendra",
"Kimberly",
"Salli",
"Joey",
"Justin",
"Kevin",
"Matthew",
"Ruth",
"Stephen"
],
"title": "TextToSpeechVoices",
"type": "string"
}
},
"description": "Text To speech node input.",
"properties": {
"voice": {
"$ref": "#/$defs/TextToSpeechVoices",
"description": "Voice to use for text to speech",
"enum_options": {
"Alloy": {
"name": "Alloy",
"provider": "OpenAI",
"voice_id": "alloy"
},
"Andrew_Neural": {
"gender": "male",
"languages": [
"en"
],
"name": "Andrew Neural",
"provider": "Azure",
"voice_id": "en-US-AndrewNeural"
},
"Ava_Neural": {
"gender": "female",
"languages": [
"en"
],
"name": "Ava Neural",
"provider": "Azure",
"voice_id": "en-US-AvaNeural"
},
"Brian_Neural": {
"gender": "male",
"languages": [
"en"
],
"name": "Brian Neural",
"provider": "Azure",
"voice_id": "en-US-BrianNeural"
},
"Danielle": {
"gender": "female",
"languages": [
"en"
],
"name": "Danielle",
"provider": "AWSPolly",
"voice_id": "Danielle"
},
"Echo": {
"name": "Echo",
"provider": "OpenAI",
"voice_id": "echo"
},
"Emma_Neural": {
"gender": "female",
"languages": [
"en"
],
"name": "Emma Neural",
"provider": "Azure",
"voice_id": "en-US-EmmaNeural"
},
"Fable": {
"name": "Fable",
"provider": "OpenAI",
"voice_id": "fable"
},
"Gregory": {
"gender": "male",
"languages": [
"en"
],
"name": "Gregory",
"provider": "AWSPolly",
"voice_id": "Gregory"
},
"Hoai_My_Neural": {
"gender": "female",
"languages": [
"vi"
],
"name": "Hoai My Neural",
"provider": "Azure",
"voice_id": "vi-VN-HoaiMyNeural"
},
"Ivy": {
"gender": "female",
"languages": [
"en"
],
"name": "Ivy",
"provider": "AWSPolly",
"voice_id": "Ivy"
},
"Joanna": {
"gender": "female",
"languages": [
"en"
],
"name": "Joanna",
"provider": "AWSPolly",
"voice_id": "Joanna"
},
"Joey": {
"gender": "male",
"languages": [
"en"
],
"name": "Joey",
"provider": "AWSPolly",
"voice_id": "Joey"
},
"Justin": {
"gender": "male",
"languages": [
"en"
],
"name": "Justin",
"provider": "AWSPolly",
"voice_id": "Justin"
},
"Kendra": {
"gender": "female",
"languages": [
"en"
],
"name": "Kendra",
"provider": "AWSPolly",
"voice_id": "Kendra"
},
"Kevin": {
"gender": "male",
"languages": [
"en"
],
"name": "Kevin",
"provider": "AWSPolly",
"voice_id": "Kevin"
},
"Kimberly": {
"gender": "female",
"languages": [
"en"
],
"name": "Kimberly",
"provider": "AWSPolly",
"voice_id": "Kimberly"
},
"Matthew": {
"gender": "male",
"languages": [
"en"
],
"name": "Matthew",
"provider": "AWSPolly",
"voice_id": "Matthew"
},
"Nam_Minh_Neural": {
"gender": "male",
"languages": [
"vi"
],
"name": "Nam Minh Neural",
"provider": "Azure",
"voice_id": "vi-VN-NamMinhNeural"
},
"Nova": {
"name": "Nova",
"provider": "OpenAI",
"voice_id": "nova"
},
"Onyx": {
"name": "Onyx",
"provider": "OpenAI",
"voice_id": "onyx"
},
"Ruth": {
"gender": "female",
"languages": [
"en"
],
"name": "Ruth",
"provider": "AWSPolly",
"voice_id": "Ruth"
},
"Salli": {
"gender": "female",
"languages": [
"en"
],
"name": "Salli",
"provider": "AWSPolly",
"voice_id": "Salli"
},
"Shimmer": {
"name": "Shimmer",
"provider": "OpenAI",
"voice_id": "shimmer"
},
"Stephen": {
"gender": "male",
"languages": [
"en"
],
"name": "Stephen",
"provider": "AWSPolly",
"voice_id": "Stephen"
},
"en_US_Casual_K": {
"languages": [
"en"
],
"name": "en-US-Casual-K",
"provider": "Google",
"voice_id": "en-US-Casual-K"
},
"en_US_Journey_D": {
"languages": [
"en"
],
"name": "en-US-Journey-D",
"provider": "Google",
"voice_id": "en-US-Journey-D"
},
"en_US_Journey_F": {
"languages": [
"en"
],
"name": "en-US-Journey-F",
"provider": "Google",
"voice_id": "en-US-Journey-F"
},
"en_US_Journey_O": {
"languages": [
"en"
],
"name": "en-US-Journey-O",
"provider": "Google",
"voice_id": "en-US-Journey-O"
},
"en_US_Neural2_A": {
"languages": [
"en"
],
"name": "en-US-Neural2-A",
"provider": "Google",
"voice_id": "en-US-Neural2-A"
},
"en_US_Neural2_C": {
"languages": [
"en"
],
"name": "en-US-Neural2-C",
"provider": "Google",
"voice_id": "en-US-Neural2-C"
},
"en_US_Neural2_D": {
"languages": [
"en"
],
"name": "en-US-Neural2-D",
"provider": "Google",
"voice_id": "en-US-Neural2-D"
},
"en_US_Neural2_E": {
"languages": [
"en"
],
"name": "en-US-Neural2-E",
"provider": "Google",
"voice_id": "en-US-Neural2-E"
},
"en_US_Neural2_F": {
"languages": [
"en"
],
"name": "en-US-Neural2-F",
"provider": "Google",
"voice_id": "en-US-Neural2-F"
},
"en_US_Neural2_G": {
"languages": [
"en"
],
"name": "en-US-Neural2-G",
"provider": "Google",
"voice_id": "en-US-Neural2-G"
},
"en_US_Neural2_H": {
"languages": [
"en"
],
"name": "en-US-Neural2-H",
"provider": "Google",
"voice_id": "en-US-Neural2-H"
},
"en_US_Neural2_I": {
"languages": [
"en"
],
"name": "en-US-Neural2-I",
"provider": "Google",
"voice_id": "en-US-Neural2-I"
},
"en_US_Neural2_J": {
"languages": [
"en"
],
"name": "en-US-Neural2-J",
"provider": "Google",
"voice_id": "en-US-Neural2-J"
},
"en_US_News_K": {
"languages": [
"en"
],
"name": "en-US-News-K",
"provider": "Google",
"voice_id": "en-US-News-K"
},
"en_US_News_L": {
"languages": [
"en"
],
"name": "en-US-News-L",
"provider": "Google",
"voice_id": "en-US-News-L"
},
"en_US_News_N": {
"languages": [
"en"
],
"name": "en-US-News-N",
"provider": "Google",
"voice_id": "en-US-News-N"
},
"en_US_Polyglot_1": {
"languages": [
"en"
],
"name": "en-US-Polyglot-1",
"provider": "Google",
"voice_id": "en-US-Polyglot-1"
},
"en_US_Standard_A": {
"languages": [
"en"
],
"name": "en-US-Standard-A",
"provider": "Google",
"voice_id": "en-US-Standard-A"
},
"en_US_Standard_B": {
"languages": [
"en"
],
"name": "en-US-Standard-B",
"provider": "Google",
"voice_id": "en-US-Standard-B"
},
"en_US_Standard_C": {
"languages": [
"en"
],
"name": "en-US-Standard-C",
"provider": "Google",
"voice_id": "en-US-Standard-C"
},
"en_US_Standard_D": {
"languages": [
"en"
],
"name": "en-US-Standard-D",
"provider": "Google",
"voice_id": "en-US-Standard-D"
},
"en_US_Standard_E": {
"languages": [
"en"
],
"name": "en-US-Standard-E",
"provider": "Google",
"voice_id": "en-US-Standard-E"
},
"en_US_Standard_F": {
"languages": [
"en"
],
"name": "en-US-Standard-F",
"provider": "Google",
"voice_id": "en-US-Standard-F"
},
"en_US_Standard_G": {
"languages": [
"en"
],
"name": "en-US-Standard-G",
"provider": "Google",
"voice_id": "en-US-Standard-G"
},
"en_US_Standard_H": {
"languages": [
"en"
],
"name": "en-US-Standard-H",
"provider": "Google",
"voice_id": "en-US-Standard-H"
},
"en_US_Standard_I": {
"languages": [
"en"
],
"name": "en-US-Standard-I",
"provider": "Google",
"voice_id": "en-US-Standard-I"
},
"en_US_Standard_J": {
"languages": [
"en"
],
"name": "en-US-Standard-J",
"provider": "Google",
"voice_id": "en-US-Standard-J"
},
"en_US_Studio_O": {
"languages": [
"en"
],
"name": "en-US-Studio-O",
"provider": "Google",
"voice_id": "en-US-Studio-O"
},
"en_US_Studio_Q": {
"languages": [
"en"
],
"name": "en-US-Studio-Q",
"provider": "Google",
"voice_id": "en-US-Studio-Q"
},
"en_US_Wavenet-G": {
"languages": [
"en"
],
"name": "en-US-Wavenet-G",
"provider": "Google",
"voice_id": "en-US-Wavenet-G"
},
"en_US_Wavenet-H": {
"languages": [
"en"
],
"name": "en-US-Wavenet-H",
"provider": "Google",
"voice_id": "en-US-Wavenet-H"
},
"en_US_Wavenet-I": {
"languages": [
"en"
],
"name": "en-US-Wavenet-I",
"provider": "Google",
"voice_id": "en-US-Wavenet-I"
},
"en_US_Wavenet-J": {
"languages": [
"en"
],
"name": "en-US-Wavenet-J",
"provider": "Google",
"voice_id": "en-US-Wavenet-J"
},
"en_US_Wavenet_A": {
"languages": [
"en"
],
"name": "en-US-Wavenet-A",
"provider": "Google",
"voice_id": "en-US-Wavenet-A"
},
"en_US_Wavenet_B": {
"languages": [
"en"
],
"name": "en-US-Wavenet-B",
"provider": "Google",
"voice_id": "en-US-Wavenet-B"
},
"en_US_Wavenet_C": {
"languages": [
"en"
],
"name": "en-US-Wavenet-C",
"provider": "Google",
"voice_id": "en-US-Wavenet-C"
},
"en_US_Wavenet_D": {
"languages": [
"en"
],
"name": "en-US-Wavenet-D",
"provider": "Google",
"voice_id": "en-US-Wavenet-D"
},
"en_US_Wavenet_E": {
"languages": [
"en"
],
"name": "en-US-Wavenet-E",
"provider": "Google",
"voice_id": "en-US-Wavenet-E"
},
"en_US_Wavenet_F": {
"languages": [
"en"
],
"name": "en-US-Wavenet-F",
"provider": "Google",
"voice_id": "en-US-Wavenet-F"
},
"vi_VN_Neural2_A": {
"languages": [
"vi"
],
"name": "vi-VN-Neural2-A",
"provider": "Google",
"voice_id": "vi-VN-Neural2-A"
},
"vi_VN_Neural2_D": {
"languages": [
"vi"
],
"name": "vi-VN-Neural2-D",
"provider": "Google",
"voice_id": "vi-VN-Neural2-D"
},
"vi_VN_Standard_A": {
"languages": [
"vi"
],
"name": "vi-VN-Standard-A",
"provider": "Google",
"voice_id": "vi-VN-Standard-A"
},
"vi_VN_Standard_B": {
"languages": [
"vi"
],
"name": "vi-VN-Standard-B",
"provider": "Google",
"voice_id": "vi-VN-Standard-B"
},
"vi_VN_Standard_C": {
"languages": [
"vi"
],
"name": "vi-VN-Standard-C",
"provider": "Google",
"voice_id": "vi-VN-Standard-C"
},
"vi_VN_Standard_D": {
"languages": [
"vi"
],
"name": "vi-VN-Standard-D",
"provider": "Google",
"voice_id": "vi-VN-Standard-D"
},
"vi_VN_Wavenet_A": {
"languages": [
"vi"
],
"name": "vi-VN-Wavenet-A",
"provider": "Google",
"voice_id": "vi-VN-Wavenet-A"
},
"vi_VN_Wavenet_B": {
"languages": [
"vi"
],
"name": "vi-VN-Wavenet-B",
"provider": "Google",
"voice_id": "vi-VN-Wavenet-B"
},
"vi_VN_Wavenet_C": {
"languages": [
"vi"
],
"name": "vi-VN-Wavenet-C",
"provider": "Google",
"voice_id": "vi-VN-Wavenet-C"
},
"vi_VN_Wavenet_D": {
"languages": [
"vi"
],
"name": "vi-VN-Wavenet-D",
"provider": "Google",
"voice_id": "vi-VN-Wavenet-D"
}
},
"title": "Voice"
},
"text": {
"description": "Text to convert to speech",
"title": "Text",
"type": "string"
}
},
"required": [
"voice",
"text"
],
"title": "TextToSpeechNodeInput",
"type": "object"
}
```
## Output Parameters
| Name | Type | Description |
| ------------ | ------ | ------------------------------------------ |
| voice\_url | string | URL to the generated audio file |
| caption\_url | string | URL to the caption/transcript file |
| duration | number | Duration of the generated audio in seconds |
View JSON Schema
```json
{
"description": "Call other workflow node output.",
"properties": {
"voice_url": {
"title": "Audio URL",
"type": "string"
},
"caption_url": {
"title": "Caption URL",
"type": "string"
},
"duration": {
"title": "Duration",
"type": "number"
}
},
"required": [
"voice_url",
"caption_url",
"duration"
],
"title": "TextToSpeechNodeOutput",
"type": "object"
}
```
## How It Works
This node converts text into natural-sounding speech audio using PixelML's text-to-speech API, which integrates multiple voice providers including Azure, OpenAI, Google, and AWS Polly. You provide the text content and select a voice from the available options, and the node generates an audio file, returns the URL to access it, along with caption data and duration information for downstream processing.
## Usage Examples
### Example 1: Convert Article to Audio with Female Voice
**Input:**
```
voice: "Joanna"
text: "Welcome to our podcast. Today we're discussing the future of artificial intelligence and its impact on everyday life."
```
**Output:**
```
voice_url: "https://storage.pixelml.com/audio/abc123.mp3"
caption_url: "https://storage.pixelml.com/captions/abc123.vtt"
duration: 12.5
```
### Example 2: Generate Vietnamese Narration
**Input:**
```
voice: "Hoai_My_Neural"
text: "Chào mừng bạn đến với hướng dẫn sử dụng sản phẩm của chúng tôi."
```
**Output:**
```
voice_url: "https://storage.pixelml.com/audio/def456.mp3"
caption_url: "https://storage.pixelml.com/captions/def456.vtt"
duration: 8.3
```
### Example 3: Create News-Style Audio with Professional Voice
**Input:**
```
voice: "en_US_News_K"
text: "Breaking news: Scientists have made a groundbreaking discovery in renewable energy technology that could revolutionize power generation worldwide."
```
**Output:**
```
voice_url: "https://storage.pixelml.com/audio/ghi789.mp3"
caption_url: "https://storage.pixelml.com/captions/ghi789.vtt"
duration: 15.7
```
## Common Use Cases
* **Podcast Generation**: Convert written content, blog posts, or articles into audio format for podcast distribution
* **Accessibility**: Create audio versions of written content to make it accessible to visually impaired users
* **E-Learning**: Generate narration for educational videos, courses, and training materials
* **Voice Notifications**: Create custom voice alerts and notifications for applications and systems
* **Audiobook Production**: Convert written books or documents into audiobook format
* **Video Voiceovers**: Generate professional voiceovers for marketing videos, explainer videos, and presentations
* **Multilingual Content**: Produce audio content in multiple languages using native-sounding voices
## Error Handling
| Error Type | Cause | Solution |
| ------------------ | ----------------------------------------------- | --------------------------------------------------------------------------------------- |
| Invalid Connection | PixelML connection is not configured or expired | Verify and update your PixelML connection credentials in the connection settings |
| Text Too Long | Input text exceeds maximum character limit | Split the text into smaller chunks and process separately, then concatenate audio files |
| Invalid Voice | Selected voice is not available or unsupported | Choose a valid voice from the available options in the voice parameter |
| API Rate Limit | Too many requests sent in a short time period | Implement delays between requests or upgrade your PixelML plan for higher limits |
| Empty Text | No text provided or text is empty | Ensure the text parameter contains at least one character |
| Network Timeout | Request timed out due to network issues | Retry the request or check network connectivity |
| Quota Exceeded | PixelML account quota has been exceeded | Upgrade your PixelML plan or wait for quota reset |
## Notes
* **Voice Selection**: Different voices are optimized for different use cases. News voices work well for formal content, Neural voices for natural conversation, and Wavenet for high-quality output.
* **Text Length**: Longer texts will result in longer processing times. Consider splitting very long texts into manageable chunks.
* **Language Support**: Ensure you select a voice that matches the language of your text for best pronunciation and natural sound.
* **Audio Format**: The generated audio is typically in MP3 format, which is widely compatible with most platforms and devices.
* **Provider Differences**: Each provider (Azure, OpenAI, Google, AWS) has distinct voice characteristics. Test multiple voices to find the best fit for your content.
* **Duration Accuracy**: The returned duration value is accurate and useful for synchronizing audio with video or other media.
* **Caption Files**: The caption\_url provides a text transcript file that can be used for subtitles or accessibility purposes.
* **Caching**: Audio files are temporarily stored and accessible via the returned URLs. Download and store files if long-term access is needed.