Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Converts text to speech using ComfyUI's Qwen-TTS service via a shell script. Supports configurable voice characters (Girl, Boy), speaking styles (Emotional, Neutral), and three model sizes (0.5B, 1.7B, 3B). Output is a WAV file saved to ComfyUI's configured output directory.
Runs entirely on a local ComfyUI instance, so there are no per-request cloud API costs or data sent to third-party services.
When to use it
- Generating voiceover narration for a video project
- Converting a written article to audio for listening while commuting
- Creating character voices for an interactive story or game prototype
- Producing audio prompts for a kiosk or accessibility interface
- Batch-generating speech samples in Mandarin Chinese for language learning content
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Generate speech from 'Welcome to the demo' using a calm girl voice and the 1.7B model
- 1Constructs a ComfyUI workflow JSON with the input text, character set to Girl, style set to Neutral, and model set to 1.7B
- 2Submits the workflow to the local ComfyUI /prompt endpoint
- 3Polls /history until the job status shows completion
- 4Retrieves the output file path from the job history response
Path to the generated WAV file in ComfyUI's output directory
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.