Hedra Plugin for OpenClaw
AI video and image generation plugin for OpenClaw powered by Hedra. Generate talking avatars, videos from text prompts, and images — all from your agent.
Features
- Image generation — create images from text prompts with configurable aspect ratios and resolutions
- Video generation — text-to-video and image-to-video with optional start/end keyframes
- Talking avatars — lip-synced avatar videos from a portrait image + audio or text-to-speech
- Voice listing — browse available TTS voices for avatar generation
- Model listing — discover available Hedra models and their capabilities
- Credit tracking — check your Hedra credit balance
Requirements
- Hedra Creator plan or above
- Hedra API key from https://www.hedra.com/api-profile
Configuration
Set your API key in openclaw.json:
{
"plugins": {
"entries": {
"hedra": {
"enabled": true,
"config": {
"apiKey": "sk_h_your-key-here"
}
}
}
}
}Or use the HEDRA_API_KEY environment variable.
| Setting | Default | Description |
|---|---|---|
apiKey | Hedra API key (falls back to HEDRA_API_KEY env var) | |
defaultVideoModel | Default model for video generation | |
defaultImageModel | Default model for image generation |
Registered Tools
| Tool | Description |
|---|---|
hedra_generate_image | Generate images from text prompts |
hedra_generate_video | Generate videos from text prompts, optionally with a start image |
hedra_generate_avatar | Generate talking avatar videos from a portrait + audio/TTS |
hedra_generation_status | Check generation status and get result URLs |
hedra_list_models | List available Hedra AI models |
hedra_list_voices | List available TTS voices for avatars |
hedra_credits | Check credit balance |
Usage Examples
Generate an image
"Generate a photo-realistic image of a mountain lake at sunset"
The agent calls hedra_list_models to choose the best image model, then hedra_generate_image with the prompt and returns the result URL.
Generate a video
"Create a 5 second video of ocean waves crashing on a beach"
The agent calls hedra_list_models to choose the best video model, then hedra_generate_video with the prompt and duration.
Create a talking avatar
"Make this portrait say 'Hello, welcome to our product demo'"
The agent calls hedra_list_voices to pick a voice, then hedra_generate_avatar with the portrait asset ID, TTS text, and voice ID.