MCP server de muapi-cli

muapi CLI

Official command-line interface for muapi.ai — generate images, videos, and audio directly from your terminal.

Agent-first design — every command works for both humans (colored output, tables) and AI agents (--output-json, --jq filtering, semantic exit codes, MCP server mode).

Related Projects

• MuAPI CLI docs — Command reference and setup guidance.
• MuAPI API reference — Underlying endpoints and prediction lifecycle.
• midjourney-api — Python SDK for Midjourney V7, V8, and Niji image generation.
• suno-api — Python SDK for Suno music, audio, and voice workflows.
• minimax-music-3-api — Python SDK for MiniMax Music 3.0 text-to-music generation on Muapi.
• awesome-minimax-music-3-prompts — Curated song prompts and lyrics-formatting guide for MiniMax Music 3.0.
• MiniMax-H3-API — Python SDK for MiniMax H3 video-generation workflows on Muapi.
• Seedance-2-Spicy-API — Python SDK and MCP server for the relaxed-moderation Seedance 2 Spicy / 2 Mini Spicy tier.
• Seedance-2.5-Spicy-API — Python SDK and MCP server for the relaxed-moderation Seedance 2.5 Spicy tier.
• Wan-3.0-Spicy-API — Python SDK and MCP server for the relaxed-moderation Wan 3.0 Spicy tier.
• awesome-minimax-h3-prompts — Prompt gallery with runnable MiniMax H3 examples for the CLI.
• Wan-3.0-API — Python SDK and MCP server for Wan 3.0-compatible AI video generation.
• Wan-3.0-Prime-API — Python SDK and MCP server for the higher-fidelity Wan 3.0 Prime tier.
• Gemini-Omni-1.1-Flash-API — Python SDK and MCP server for Google's newly announced Gemini Omni 1.1 Flash update.
• Awesome-Gemini-Omni-API-Prompts — Curated Gemini Omni prompts and REST examples.
• Open-Generative-AI — Browser-based GUI for the same models — no CLI required
• Awesome-GPT-Image-2-API-Prompts — Curated prompt library to run via this CLI
• ai-creator-academy — free curriculum teaching creators how to monetize generative AI, including running models via this CLI
• Flux-3-Dev-API — Python wrapper for Black Forest Labs' FLUX 3 (Dev variant) — text-to-image, image-to-image, text-to-video, image-to-video
• Grok-Imagine-Image-2-API — Python SDK and MCP server for Grok Imagine Image 2.0 generation and editing through MuAPI
• awesome-flux-3-api-prompts — FLUX 3 API guide, prompts, and parameters
• flux-3-video-api — Python wrapper focused on FLUX 3 Text-to-Video and Image-to-Video
• seedance-2.5-mcp — MCP server for generating Seedance 2.5 Preview videos through MuAPI.
• seedance-2-mcp — MCP server for generating Seedance 2 videos through MuAPI.
• Text-to-Speech-API — speech model comparison and raw API examples that can be run through the CLI.
• Speech-to-Text-API — transcription and audio-analysis examples for the same Muapi endpoints.
• Image-Enhancement-API — image upscaling and background-removal API examples.
• Video-Utilities-API — video enhancement and MMAudio utility examples.
• AI-3D-Model-API — Tripo3D and Meshy API comparisons and examples.
• muapi-skills — Agent skills (npx skills add SamurAIGPT/muapi-skills) that let Claude Code, Cursor and Codex discover, price and run Muapi models through this CLI.

Install

# npm (recommended — no Python required)
npm install -g muapi-cli

# pip
pip install muapi-cli

# or run without installing
npx muapi-cli --help

Quick Start

# New user? Create an account
muapi auth register --email you@example.com --password "..."
muapi auth verify --email you@example.com --otp 123456
muapi auth login --email you@example.com --password "..."

# Or paste an existing API key
muapi auth configure --api-key "YOUR_KEY"

# Generate — pick a curated verb…
muapi image generate "a cyberpunk city at night" --model flux-dev
muapi video generate "a dog running on a beach" --model kling-master
muapi audio create "upbeat lo-fi hip hop for studying"

# …or run any model by endpoint name (schema-driven, covers the whole catalog)
muapi run flux-dev-image -p "a cyberpunk city at night"
muapi run seedance-2-text-to-video -p "drone shot over snowy peaks" -i duration=5
muapi run <model> -h          # introspects the live OpenAPI schema

# Check balance
muapi account balance

# Wait for an existing job
muapi predict wait <request_id>

Commands

muapi auth

Command	Description
muapi auth register --email x --password y	Create a new account (sends OTP)
muapi auth verify --email x --otp 123456	Verify email after registration
muapi auth login --email x --password y	Log in and save API key automatically
muapi auth forgot-password --email x	Send password reset OTP
muapi auth reset-password --email x --otp y --password z	Reset password
muapi auth configure	Manually save an API key
muapi auth whoami	Show current API key (masked)
muapi auth logout	Remove stored API key

muapi account

Command	Description
muapi account balance	Show current credit balance
muapi account topup --amount 20	Add credits via Stripe checkout

muapi keys

Command	Description
muapi keys list	List all API keys on your account
muapi keys create --name label	Create a new API key (shown once)
muapi keys delete <id>	Delete an API key by ID

muapi image

Command	Description
muapi image generate <prompt>	Text-to-image generation
muapi image edit <prompt> --image <url>	Image-to-image editing
muapi image models	List available models

Models: flux-dev, flux-schnell, flux-kontext-dev/pro/max, hidream-fast/dev/full, wan2.1, reve, gpt4o, midjourney, seedream, qwen

muapi video

Command	Description
muapi video generate <prompt>	Text-to-video generation
muapi video from-image <prompt> --image <url>	Image-to-video animation
muapi video models	List available models

Models: veo3, veo3-fast, kling-master, kling-std, kling-pro, wan2.1/2.2, seedance-pro/lite, hunyuan, runway, pixverse, vidu, minimax-std/pro

muapi audio

Command	Description
muapi audio create <prompt>	Create music with Suno
muapi audio remix <song-id>	Remix an existing Suno song
muapi audio extend <song-id>	Extend a Suno song
muapi audio from-text <prompt>	Generate audio with MMAudio
muapi audio from-video <video-url>	Add AI audio to a video

muapi enhance

Command	Description
muapi enhance upscale <url>	AI image upscaling
muapi enhance bg-remove <url>	Remove background
muapi enhance face-swap --source <url> --target <url>	Face swap image/video
muapi enhance skin <url>	Skin enhancement
muapi enhance colorize <url>	Colorize B&W photo
muapi enhance ghibli <url>	Ghibli anime style
muapi enhance anime <url>	Anime style conversion
muapi enhance extend <url>	Outpaint/extend image
muapi enhance product-shot <url>	Professional product photo
muapi enhance erase <url> --mask <url>	Object removal

muapi edit

Command	Description
muapi edit effects --video <url> --effect <name>	AI video/image effects
muapi edit lipsync --video <url> --audio <url>	Lip sync to audio
muapi edit dance --image <url> --video <url>	Make person dance
muapi edit dress --image <url>	Change clothing
muapi edit clipping <video-url>	AI highlight extraction

muapi run — generic, schema-driven runner

Reaches any model in the muapi.ai catalog by endpoint name, even ones not covered by the curated image / video / audio / enhance / edit verbs. The input schema is fetched from the live OpenAPI spec, so muapi run <model> -h always reflects the real, current parameters.

Command	Description
muapi run <model> -h	Print model-specific inputs from the live OpenAPI schema
muapi run <model> -p "..."	Run with a prompt
muapi run <model> -p "..." -i k=v -i k=v	Pass arbitrary inputs (JSON-parsed when valid)
muapi run <model> --input-file inputs.json	Inputs from a JSON file
muapi run <model> ... --dry-run	Show the request body without sending

<model> accepts either a real endpoint slug (flux-dev-image, nano-banana-2, seedance-2-text-to-video) or a short alias from the curated tables (flux-dev, seedream, kling-master).

Merge order for inputs (later wins): --input-file → -i k=v → -p prompt.

# Discover a model's real inputs
muapi run nano-banana-2 -h

# Run it
muapi run nano-banana-2 -p "a logo for a coffee shop" -i num_images=2 --download ./out

# Pipe-safe JSON
muapi run flux-dev-image -p "..." --output-json --jq '.outputs[0]'

muapi predict

Command	Description
muapi predict result <id>	Fetch current status (no polling)
muapi predict wait <id>	Wait until complete

muapi upload

Command	Description
muapi upload file <path>	Upload a local file → get hosted URL

muapi models

Command	Description
muapi models list	List all models
muapi models list --category video	Filter by category

muapi config

Command	Description
muapi config set <key> <value>	Set a persistent default
muapi config get <key>	Read a config value
muapi config list	Show all config

Useful keys: output (json/human), model.image, model.video, no_color (true/false)

muapi docs

Command	Description
muapi docs openapi	Fetch the full OpenAPI spec
muapi docs open	Open Swagger UI in browser

muapi mcp

Command	Description
muapi mcp serve	Start MCP server (stdio) for AI agents

Global Options

Flag	Description
--wait / --no-wait	Poll until done (default: --wait)
--output-json / -j	Print raw JSON response
--jq <expr>	Filter JSON output (e.g. '.outputs[0]')
--download <dir> / -d	Auto-download outputs to directory
--no-color	Disable colored output

Environment Variables

Variable	Description
MUAPI_API_KEY	API key (overrides keychain/config)
MUAPI_BASE_URL	Override API base URL
NO_COLOR	Disable colored output

Exit Codes

Code	Meaning
0	Success
1	General error
3	Authentication error
4	Rate limited
5	Not found
6	Billing error
7	Timeout
8	Validation error

MCP Server

muapi supports two MCP transport modes. Both expose the same 19 tools: image generate/edit, video generate/from-image, audio create/from-text, enhance (upscale/bg-remove/face-swap/ghibli), edit lipsync/clipping, predict result, upload file, keys list/create/delete, account balance/topup.

Option 1 — Hosted (Recommended, no CLI required)

The hosted MCP server at https://api.muapi.ai/mcp uses the standard Streamable HTTP transport. Any MCP client can connect with just your API key — no CLI install needed.

Claude Code:

claude mcp add --transport http muapi \
  https://api.muapi.ai/mcp \
  --header "Authorization: Bearer YOUR_MUAPI_KEY"

Cursor — add to mcp.json (Cmd+Shift+P → Open MCP settings):

{
  "mcpServers": {
    "muapi": {
      "url": "https://api.muapi.ai/mcp",
      "headers": { "Authorization": "Bearer YOUR_MUAPI_KEY" }
    }
  }
}

Windsurf — open Settings → MCP:

{
  "mcpServers": {
    "muapi": {
      "serverUrl": "https://api.muapi.ai/mcp",
      "headers": { "Authorization": "Bearer YOUR_MUAPI_KEY" }
    }
  }
}

Option 2 — stdio via CLI (Claude Desktop)

Run the CLI as a local stdio MCP server. Requires the CLI to be installed.

muapi mcp serve

Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "muapi": {
      "command": "muapi",
      "args": ["mcp", "serve"],
      "env": { "MUAPI_API_KEY": "your-key-here" }
    }
  }
}

Any stdio-compatible client — use the command + args pattern above, substituting your client's config format.

For more details on the self-hosted MCP server see muapi-mcp-server.

Grok Build Plugin

This repository also ships a Grok Build plugin that connects Grok to Muapi's hosted MCP server. Install the muapi plugin from the Grok Build marketplace, then set MUAPI_API_KEY in the environment used by Grok Build:

export MUAPI_API_KEY="YOUR_MUAPI_KEY"

The plugin provides live model discovery plus image, video, audio, enhancement, editing, and async-result tools. Generation and enhancement consume Muapi credits; account top-ups, key management, and external publishing remain approval-gated by the plugin instructions.

For a direct project-local setup without waiting for marketplace review, add the hosted MCP server to Grok Build with:

grok mcp add --transport http muapi \
  https://api.muapi.ai/mcp \
  --header "Authorization: Bearer ${MUAPI_API_KEY}"

Agentic Pipeline Examples

# Full onboarding without human intervention
muapi auth register --email agent@example.com --password "secret"
muapi auth verify --email agent@example.com --otp 123456
muapi auth login --email agent@example.com --password "secret"
muapi account balance --output-json
muapi account topup --amount 10 --output-json --no-open

# Submit async, capture request_id, poll when ready
REQUEST_ID=$(muapi video generate "a dog on a beach" \
  --model kling-master --no-wait --output-json --jq '.request_id' | tr -d '"')
muapi predict wait "$REQUEST_ID" --download ./outputs

# Chain: upload → edit → download
URL=$(muapi upload file ./photo.jpg --output-json --jq '.url' | tr -d '"')
muapi image edit "make it look like a painting" --image "$URL" \
  --model flux-kontext-pro --download ./outputs

# Rotate API keys programmatically
NEW_KEY=$(muapi keys create --name "ci-$(date +%Y%m%d)" --output-json --jq '.api_key' | tr -d '"')
OLD_ID=$(muapi keys list --output-json --jq '.[0].id')
muapi keys delete "$OLD_ID" --yes

# Discover available endpoints
muapi docs openapi --jq '.paths | keys[]'

Shell Completions

muapi --install-completion bash
muapi --install-completion zsh
muapi --install-completion fish