What you can do
- Use GPT-4 as a coding agent for complex tasks
- Generate images with DALL-E via natural language
- Transcribe audio files with Whisper
- Chain multiple AI models in workflows
- Run multi-step reasoning and code generation
Install the skills
Coding agent
npx clawhub@latest install coding-agent
Image generation
npx clawhub@latest install openai-image-gen
Speech to text
npx clawhub@latest install openai-whisper
The OpenAI integration spans three skills: coding-agent for GPT-powered development, openai-image-gen for DALL-E image generation, and openai-whisper for audio transcription.
Example commands
# Generate an image
openclaw ask "Generate an image of a futuristic dashboard"
β Image generated: futuristic-dashboard.png (1024x1024)
# Transcribe audio
openclaw ask "Transcribe this meeting recording"
β Transcribed: meeting-2025-03.txt (42 min, 8200 words)
# Code with GPT-4
openclaw ask "Write a REST API for user authentication"
β Generated 4 files: routes, middleware, model, tests
How it works
OpenClaw CLI acts as a watchdog daemon that keeps your AI agent always running. The OpenAI integration brings GPT-4, DALL-E, and Whisper to your CLI workflow. Your agent can write code, generate images, and transcribe audio β all through natural language. Set your OpenAI API key once, and your agent orchestrates the models.
Get OpenClaw CLI β Free
Your AI agent, always running. Zero config.
npm
npm install -g openclaw-cli
Homebrew
brew install openclaw/tap/openclaw-cli