Skip to content
OpenMicroDrama

MiniMax-AI /cli

MiniMax images, video, speech and text from your terminal
Model toolingNo licenseEnglish README
Open on GitHubSave
Stars
2.2k
30 days
New
Last push
2 days ago

About cli

mmx-cli is the official command-line tool for the MiniMax AI Platform. You can make images, video, speech and text from a shell or an AI coding agent. It needs a paid MiniMax Token Plan and API keys. The repo has no license field set, even though the README links an MIT license.

Best for: Developers and agent users who want MiniMax models inside scripts, CI jobs or agent workflows instead of a web UI.

What it does

  • Chats with text models, with streaming and JSON output
  • Makes images with aspect-ratio and batch controls
  • Generates video as a background job, then downloads it
  • Speaks with 30+ voices; transcribes audio to srt or vtt
  • Reads images and runs MiniMax web search
  • Works with Global (api.minimax.io) and CN (api.minimaxi.com) regions

Quickstart

Rewritten from the README. Check the repo for the latest steps.

  1. 1
    Add the skill to your agent with the command below, or install the CLI with npm install -g mmx-cli (Node.js 18+).
    npx skills add MiniMax-AI/cli -y -g
  2. 2
    Log in with mmx auth login, or skip the prompts with the command below.
    mmx auth login --api-key sk-xxxxx
  3. 3
    Make media, for example the command below or the next command.
    mmx image "A cat in a spacesuit"
    mmx video generate --prompt "Ocean waves at sunset" --download sunset.mp4
  4. 4
    Do speech with the command below and the next command.
    mmx speech synthesize --text "Hello!" --out hello.mp3
    mmx speech transcribe --file meeting.mp3
  5. 5
    Check your usage with mmx quota and your settings with mmx config show.

Models and languages

Models it works with

  • MiniMax H3
  • Hailuo-2.3
  • MiniMax-M3
  • asr-1.0

Interface and docs

  • English
  • Chinese

Alternatives

Other model tooling projects people compare with cli.

letorig

video-generator-client

Puts Seedance, Kling, MiniMax and Wan video models behind one Python tool

Model tooling
  • Seedance 2.5
  • Seedance 2.0
  • Kling 3.0
  • +7
551 GitHub stars4 days agoMIT

MiniMax-AI

MiniMax-MCP

Plugs MiniMax speech, voice cloning, image and video tools into your MCP client

Model tooling
  • MiniMax-Hailuo-02
1.6k GitHub stars1 month agoMIT

higgsfield-ai

cli

40+ Higgsfield image, video, 3D and audio models from your terminal

Model tooling
  • Nano Banana Pro
  • Nano Banana 2 Lite
  • Gemini Omni Flash
  • +7
621 GitHub stars12 days agoMIT

Wan-Video

Wan2.2

Open video models for text, image, speech and character animation

Model tooling
  • Wan 2.2-T2V-A14B
  • Wan 2.2-I2V-A14B
  • Wan 2.2-TI2V-5B
  • +5
18k GitHub stars10 days agoApache-2.0

Featured on OpenMicroDrama

Do you maintain cli? Add this badge to your README so English speakers can find our write-up.

Featured on OpenMicroDrama
markdown
[![Featured on OpenMicroDrama](https://openmicrodrama.com/badges/featured.svg)](https://openmicrodrama.com/projects/minimax-ai-cli)

The badge links to this page. Want your description changed? Email support@openmicrodrama.com.

Reviewed Oct 1, 2026. Official MiniMax tool. The README links an MIT license, but there's no LICENSE file in the repo yet. Needs a paid MiniMax plan. We write these descriptions ourselves. The repo's own docs are the final word.