zhipu-tts

Original source / Raw SKILL.md

---
name: zhipu-tts
description: Text-to-speech conversion using Zhipu AI (BigModel) GLM-TTS model. Use when you need to convert text to audio files with various voice options. Supports Chinese text synthesis with multiple voice personas, speed control, and output formats.
metadata:
  {
    "openclaw":
      {
        "requires": { "bins": ["jq"], "env": ["ZHIPU_API_KEY"] },
      },
  }
---

# Zhipu AI Text-to-Speech

Convert Chinese text to natural-sounding speech using Zhipu AI's GLM-TTS model.

## Setup

**1. Get your API Key:**
Get a key from [Zhipu AI Console](https://bigmodel.cn/usercenter/proj-mgmt/apikeys)

**2. Set it in your environment:**
```bash
export ZHIPU_API_KEY="your-key-here"
```

## Available Voices

### System Voices (Pre-built)

- **tongtong** (彤彤) - Default voice, balanced tone
- **chuichui** (锤锤) - Male voice, deeper tone
- **xiaochen** (小陈) - Young professional voice
- **jam** - 动动动物圈 Jam voice
- **kazi** - 动动动物圈 Kazi voice
- **douji** - 动动动物圈 Douji voice
- **luodo** - 动动动物圈 Luodo voice

## Usage

### Basic Text-to-Speech

Convert text to speech with default settings (tongtong voice, normal speed, WAV format):

```bash
bash scripts/text_to_speech.sh "你好，今天天气怎么样"
```

### Advanced Options

Specify voice, speed, format, and output filename:

```bash
bash scripts/text_to_speech.sh "欢迎使用智能语音服务" xiaochen 1.2 wav greeting.wav
```

**Parameters:**
- `text` (required): Chinese text to convert (max 1024 characters)
- `voice` (optional): tongtong (default), chuichui, xiaochen, jam, kazi, douji, luodo
- `speed` (optional): Speech speed from 0.5 to 2.0 (default: 1.0)
- `output_format` (optional): wav (default), pcm
- `output_file` (optional): Output filename (default: output.{format})

## Voice Selection Guide

**Choose tongtong (default) for:**
- General purpose narration
- Professional presentations
- Balanced tone requirements

**Choose chuichui for:**
- Male voice needed
- Deeper, authoritative tone
- Documentary or formal content

**Choose xiaochen for:**
- Young, energetic tone
- Modern, casual content
- Friendly assistant vibe

**Choose jam/kazi/douji/luodo for:**
- Entertainment content
- Character voices
- Creative projects

## Speed Control

**Recommended speeds:**
- **0.8-1.0**: Clear, professional narration
- **1.0-1.2**: Natural conversational pace (default: 1.0)
- **1.2-1.5**: Energetic, upbeat delivery
- **1.5-2.0**: Fast-paced summaries (may reduce clarity)

## Output Formats

**WAV (recommended):**
- Standard audio format
- Widely compatible
- Better quality preservation

**PCM:**
- Raw audio format
- Smaller file size
- Requires additional processing for playback

## Examples

Create a professional greeting:

```bash
bash scripts/text_to_speech.sh "您好，感谢致电智能客服，请按1选择中文服务" tongtong 1.0 wav greeting.wav
```

Generate an energetic announcement:

```bash
bash scripts/text_to_speech.sh "热烈欢迎各位嘉宾参加今天的活动！" xiaochen 1.3 wav announcement.wav
```

Create a calm narration:

```bash
bash scripts/text_to_speech.sh "在这个宁静的夜晚，让我们一起欣赏美丽的星空" chuichui 0.9 wav narration.wav
```

## Character Limits

- Maximum input: **1024 characters** per request
- For longer texts, split into multiple segments
- Combine audio files post-generation

## Audio Quality Tips

**Best practices:**
- Use punctuation for natural pauses (commas, periods)
- Break long sentences into shorter segments
- Use appropriate line breaks for paragraph pauses
- Test speed settings for your specific content

**Sample rate:** Generated audio uses 24000 Hz sampling rate for optimal quality.

## Troubleshooting

**Text Length Issues:**
- Split texts longer than 1024 characters
- Process segments separately
- Combine using audio editing tools

**Audio Quality Issues:**
- Check text encoding (use UTF-8)
- Verify punctuation placement
- Adjust speed settings
- Try different voices

**File Playback Issues:**
- Ensure format compatibility with your player
- WAV format works on most systems
- PCM may require conversion

## API Notes

- Responses are returned as audio files
- Watermarking enabled by default (can be disabled in account settings)
- No strict rate limiting documented
- Audio generation typically completes in 1-3 seconds


---

## Referenced Files

> The following files are referenced in this skill and included for context.

### scripts/text_to_speech.sh

```bash
#!/bin/bash
# Zhipu AI Text-to-Speech Script
# Usage: ./text_to_speech.sh "text" [voice] [speed] [output_format]

set -e

# Configuration
API_ENDPOINT="https://open.bigmodel.cn/api/paas/v4/audio/speech"

# Get API key from environment
if [ -z "$ZHIPU_API_KEY" ]; then
    echo "Error: ZHIPU_API_KEY environment variable is not set" >&2
    echo "" >&2
    echo "To fix:" >&2
    echo "1. Get a key from https://bigmodel.cn/usercenter/proj-mgmt/apikeys" >&2
    echo "2. Run: export ZHIPU_API_KEY=\"your-key\"" >&2
    exit 1
fi

# Parse arguments
TEXT="$1"
VOICE="${2:-tongtong}"
SPEED="${3:-1.0}"
OUTPUT_FORMAT="${4:-wav}"
OUTPUT_FILE="${5:-output.${OUTPUT_FORMAT}}"

# Validate text
if [ -z "$TEXT" ]; then
    echo "Usage: $0 \"text\" [voice] [speed] [output_format] [output_file]" >&2
    echo "" >&2
    echo "Examples:" >&2
    echo "  $0 \"你好，今天天气怎么样\"" >&2
    echo "  $0 \"欢迎使用智能语音服务\" xiaochen 1.2 wav greeting.wav" >&2
    echo "" >&2
    echo "Voices: tongtong (default), chuichui, xiaochen, jam, kazi, douji, luodo" >&2
    echo "Speed: 0.5-2.0 (default: 1.0)" >&2
    echo "Format: wav (default), pcm" >&2
    exit 1
fi

# Build request payload
PAYLOAD=$(jq -n \
    --arg model "glm-tts" \
    --arg input "$TEXT" \
    --arg voice "$VOICE" \
    --argjson speed "$SPEED" \
    --arg response_format "$OUTPUT_FORMAT" \
    '{
        model: $model,
        input: $input,
        voice: $voice,
        speed: $speed,
        response_format: $response_format
    }')

# Make API request
echo "Converting text to speech..." >&2
echo "Voice: $VOICE, Speed: $SPEED, Format: $OUTPUT_FORMAT" >&2
echo "" >&2

RESPONSE=$(curl -s -X POST "$API_ENDPOINT" \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $ZHIPU_API_KEY" \
    -d "$PAYLOAD" \
    --output "$OUTPUT_FILE" \
    -w "%{http_code}")

# Check for errors
if [ "$RESPONSE" != "200" ]; then
    echo "Error: HTTP $RESPONSE" >&2
    if [ -f "$OUTPUT_FILE" ]; then
        cat "$OUTPUT_FILE" >&2
        rm "$OUTPUT_FILE"
    fi
    exit 1
fi

echo "Audio saved to: $OUTPUT_FILE" >&2
echo "Text: $TEXT" >&2

# Get file info
if command -v file &> /dev/null; then
    FILE_INFO=$(file "$OUTPUT_FILE")
    echo "File info: $FILE_INFO" >&2
fi

if command -v ls &> /dev/null; then
    FILE_SIZE=$(ls -lh "$OUTPUT_FILE" | awk '{print $5}')
    echo "File size: $FILE_SIZE" >&2
fi

echo "$OUTPUT_FILE"

```



---

## Skill Companion Files

> Additional files collected from the skill directory layout.

### README.md

```markdown
# Zhipu AI TTS Skill

Text-to-speech conversion using Zhipu AI (BigModel) GLM-TTS model. Convert Chinese text to natural-sounding speech with multiple voice options.

## Features

- 🎙️ **Multiple Voices**: 7 different voice personas (tongtong, chuichui, xiaochen, jam, kazi, douji, luodo)
- ⚡ **Speed Control**: Adjustable speech speed from 0.5x to 2.0x
- 🎵 **Multiple Formats**: WAV and PCM output formats
- 🇨🇳 **Chinese Language**: Optimized for Mandarin Chinese synthesis
- 📝 **Long Text Support**: Up to 1024 characters per request
- 🔊 **High Quality**: 24000 Hz sampling rate for optimal audio quality

## Requirements

- `jq` - JSON processor
- `ZHIPU_API_KEY` environment variable

## Quick Start

```bash
# Install dependencies (if needed)
sudo apt-get install jq

# Set your API key
export ZHIPU_API_KEY="your-key-here"

# Convert text to speech (default settings)
bash scripts/text_to_speech.sh "你好，今天天气怎么样"

# With custom voice and speed
bash scripts/text_to_speech.sh "欢迎使用智能语音服务" xiaochen 1.2 wav greeting.wav
```

## Available Voices

- **tongtong** (彤彤) - Default balanced tone
- **chuichui** (锤锤) - Male voice, deeper tone
- **xiaochen** (小陈) - Young professional voice
- **jam** - 动动动物圈 Jam voice
- **kazi** - 动动动物圈 Kazi voice
- **douji** - 动动动物圈 Douji voice
- **luodo** - 动动动物圈 Luodo voice

## Use Cases

- 📚 Audiobook creation
- 🎮 Game character voices
- 📢 Announcement systems
- 🤖 Virtual assistants
- 🎬 Video dubbing
- 📻 Radio content generation

## Parameters

- `text` (required): Chinese text to convert (max 1024 characters)
- `voice` (optional): Voice persona (default: tongtong)
- `speed` (optional): Speech speed 0.5-2.0 (default: 1.0)
- `output_format` (optional): wav or pcm (default: wav)
- `output_file` (optional): Output filename (default: output.{format})

## Examples

```bash
# Professional greeting
bash scripts/text_to_speech.sh "您好，感谢致电智能客服" tongtong 1.0 wav greeting.wav

# Energetic announcement
bash scripts/text_to_speech.sh "热烈欢迎各位嘉宾！" xiaochen 1.3 wav announcement.wav

# Calm narration
bash scripts/text_to_speech.sh "在这个宁静的夜晚" chuichui 0.9 wav narration.wav
```

## Author

franklu0819-lang

## License

MIT

```

### _meta.json

```json
{
  "owner": "franklu0819-lang",
  "slug": "zhipu-tts",
  "displayName": "Zhipu AI TTS",
  "latest": {
    "version": "1.0.0",
    "publishedAt": 1771684745050,
    "commit": "https://github.com/openclaw/skills/commit/5cfba21d911fd636f0708cdda4c192ac4f144e43"
  },
  "history": []
}

```