ClawHub

MiniMax 媒体生成 Unified MiniMax media generation skill for audio, image, and video creation with a single command entrypoint.MiniMax Skill 是一个统一的媒体生成技能，把文本转语音、文生图、文生视频三类能力收口为一个入口。安装后只需配置自己的 MINIMAX_API_KEY，即可通过统一命令生成音频、图片和视频，适合在 OpenClaw / Codex 类环境中直接调用。

Workflows

Unified MiniMax media generation skill for Token Plan workflows. Use when the user asks to generate audio, speech, TTS, narration, images, illustrations, posters, covers, thumbnails, videos, motion clips, text-to-video, image-to-video, or needs MiniMax API setup instructions across Linux, macOS, Windows, Docker, or CI.

Install

openclaw skills install minimax-skill

MiniMax Media Generation

Use this skill to generate MiniMax audio, images, and videos from one command surface.

Requirements

  • Python 3
  • requests package
  • MINIMAX_API_KEY environment variable

For system-specific API key setup, read references/setup.md.

Unified Command

Run scripts/minimax.py with one mode: audio, image, or video.

python scripts/minimax.py {audio|image|video} [options]

Audio

Generate an MP3 file from text:

python scripts/minimax.py audio \
  --text "你好，今天我们聊一下 MiniMax。" \
  --output /tmp/minimax-audio.mp3

Useful options:

  • --model speech-2.8-turbo for lower latency
  • --model speech-2.8-hd for higher fidelity
  • --voice-id <voice_id> to select a voice when available

Image

Generate images into a directory:

python scripts/minimax.py image \
  --prompt "一只穿西装的橘猫，电影感，柔光" \
  --output /tmp/minimax-images \
  --aspect-ratio 1:1

Useful options:

  • --model image-01
  • --model image-01-live for stronger style control when available
  • --n 2 to request multiple images
  • --image-file <url_or_base64> for image-to-image workflows when supported

Outputs are saved as output-0.jpeg, output-1.jpeg, etc.

Video

Generate a video file:

python scripts/minimax.py video \
  --prompt "镜头缓慢推进，一只橘猫走过雨夜街道，霓虹灯反光" \
  --output /tmp/minimax-video.mp4

Useful options:

  • --model MiniMax-Hailuo-2.3
  • --first-frame-image <url_or_base64> for image-to-video
  • --last-frame-image <url_or_base64> for start/end-frame video
  • --subject-reference '<json>' for subject reference workflows
  • --timeout 1800 to control polling timeout in seconds

Direct Scripts

The unified command delegates to standalone scripts in the same skill:

  • scripts/generate_audio.py
  • scripts/generate_image.py
  • scripts/generate_video.py

Use these directly only when a narrower, mode-specific command is preferred.

Safe Sharing Notes

  • Do not include a real API key in the skill package.
  • Ask users to configure MINIMAX_API_KEY locally.
  • Keep generated media outside the skill folder, for example under /tmp or a project output directory.
More in Workflows