Generate Text, Images, Video, Speech, and Music
Agent guide for the mmx CLI: generate text, images, video, speech, and music, plus web search, via the MiniMax platform.
Why it matters
Leverage the MiniMax AI platform via a command-line interface to generate a wide range of content, including text, images, video, speech, and music, and perform web searches.
Outcomes
What it gets done
Generate text content using chat completion.
Create images from textual prompts.
Produce video content based on descriptions.
Synthesize speech from text.
Compose music with or without lyrics.
Perform web searches and retrieve results.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/ag-mmx-cli | bash Overview
MiniMax CLI - Agent Skill Guide
An agent-oriented CLI guide for mmx, covering text, image, video, speech, and music generation plus web search and vision through the MiniMax platform. Use when a script or coding agent needs to generate or inspect media, text, or search results via MiniMax from the terminal rather than a web UI.
What it does
An agent skill guide for the mmx CLI, which drives the MiniMax AI platform to generate text, images, video, speech, and music, and to run web search and image-understanding (vision) queries, all from the terminal. It documents installation and authentication:
npm install -g mmx-cli
Authentication persists via mmx auth login --api-key, with OAuth saved to ~/.mmx/credentials.json and API keys to ~/.mmx/config.json, or an --api-key flag passed per call; region is auto-detected with --region global/--region cn overrides. It lists six agent-context flags recommended for CI/agent use: --non-interactive, --quiet, --output json, --async, --dry-run, and --yes. It covers seven commands with default models and worked examples: text chat (default MiniMax-M3, with --model MiniMax-M2.7 to pin the previous generation), image generate (model image-01), video generate (default MiniMax-Hailuo-2.3, async by default), speech synthesize (default speech-2.8-hd, max 10k characters), music generate (model music-2.6-free, instrumental or with auto-generated lyrics), search query, and vision describe. It also documents two piping patterns - chaining image generation into vision description, and an async video generate/poll/download workflow using jq to extract the task ID - and a ten-code exit-code table covering success, general/usage errors, authentication failure, quota exceeded, timeout, and content-filter triggers.
When to use - and when NOT to
Use it when a user wants to generate or inspect text, images, video, speech, music, web-search results, or MiniMax API resources through the mmx terminal CLI. It requires a configured MiniMax account and valid authentication before any API-backed command will work; the skill documents CLI usage only and does not replace provider policy review, content-safety checks, or downstream validation of returned file paths and URLs.
Inputs and outputs
Input is a natural-language generation request - a chat message, an image/video/music prompt, text to synthesize, or a search query - plus the target mmx subcommand. Output is the generated media or data, ideally as machine-readable JSON when --output json --quiet is used, or an async task ID for video that is polled via mmx video task get and retrieved with mmx video download.
Integrations
Wraps the MiniMax AI platform's text (MiniMax-M3/M2.7), image (image-01), video (MiniMax-Hailuo-2.3), speech (speech-2.8-hd), music (music-2.6-free), search, and vision APIs behind a single CLI, with jq used in the documented async video pipeline to parse task IDs from JSON output.
Who it's for
Developers and coding agents that need to call MiniMax's generation and search APIs from a script or CI pipeline rather than a web UI, especially in non-interactive automation contexts.
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.