mmx-cli
Generate AI media from the terminal
Create text, images, video, speech, and music using the MiniMax AI platform directly from your terminal. Avoid complex API integration by using simple command-line commands for all your AI generation needs.
Download the skill ZIP
Upload in Claude
Go to Settings → Capabilities → Skills → Upload skill
Toggle on and start using
Test it
Using "mmx-cli". mmx text chat --message "user:What is the capital of France?" --output json --quiet
Expected outcome:
The AI responds with the answer in JSON format, containing the response text and token usage information.
Using "mmx-cli". mmx image generate --prompt "A cat in a spacesuit" --output json --quiet
Expected outcome:
The command returns one or more image URLs pointing to the generated images. In quiet mode, URLs appear one per line on stdout.
Using "mmx-cli". mmx speech synthesize --text "Hello world" --out hello.mp3 --quiet
Expected outcome:
The command creates an MP3 audio file at the specified path. The file path is printed to stdout upon completion.
Security Audit
Low RiskThe static analyzer reported 202 potential issues, all of which are false positives. SKILL.md is a markdown documentation file for the mmx CLI tool. The scanner misidentified markdown code block backticks as Ruby/shell backtick execution, model version strings as weak cryptographic algorithms, and CLI command examples as system reconnaissance. The skill instructs an AI agent to run the mmx CLI with user-provided parameters, which involves legitimate external command execution, filesystem access for credentials (~/.mmx/), and environment variable usage for API keys. No malicious intent, prompt injection attempts, or data exfiltration patterns were found.
Low Risk Issues (3)
Risk Factors
⚙️ External commands (1)
📁 Filesystem access (2)
🔑 Env variables (1)
Quality Score
What You Can Build
Content creation workflow
Content creators generate images, video, speech, and music directly from terminal scripts without switching to web interfaces or writing API integration code.
AI prototyping and development
Developers quickly test MiniMax model outputs and integrate generation features into applications using the CLI as a prototyping tool.
Automated media production
Teams automate media generation pipelines by chaining mmx commands with standard Unix tools like jq for JSON processing.
Try These Prompts
Generate an image of a cat sitting on a cloud at sunset. Save it to the current directory.
Start a conversation with MiniMax using a system prompt that sets it as a Python tutor. Then ask it to explain list comprehensions with examples.
Generate a short video of ocean waves. Use async mode so I can continue working. Show me the task ID, then check its status and download when complete.
Generate an image of a sunset. Then use vision describe to analyze the generated image and describe what it contains. Pipe the results through jq to extract only the description text.
Best Practices
- Always use --non-interactive and --quiet flags in automated or CI environments to prevent unexpected prompts and progress output.
- Use --output json for machine-readable results when piping command output to other tools or scripts.
- Set your MINIMAX_API_KEY as an environment variable rather than passing it as a flag in each command to avoid exposing credentials in shell history.
Avoid
- Do not pass API keys directly in command flags for shared or logged environments. Use environment variables or the mmx auth login command instead.
- Do not use the CLI in interactive mode within automated scripts. Interactive prompts will block execution and cause timeouts.
- Do not ignore the --async flag for video generation in production workflows. Synchronous video generation can block for extended periods.