WaveSpeed MCP

npx add-mcp 'https://api.wavespeed.ai/mcp'

Generate images and videos from Claude Code, Codex and your terminal

One open-source CLI to run any WaveSpeed AI model. Image, video, audio, 3D. Built for terminals and coding agents like Codex, Claude Code, Claude Cowork, and Cursor.

CLI

@wavespeed/cli

Install once, run any model from your shell or scripts.

$npm install -g @wavespeed/cli

Set up WaveSpeed in your agent

Pick your tool for the quickest setup. Every option gets the full model catalog, with prices before you spend.

Claude Code

Plugin

Adds the hosted MCP server and the WaveSpeed skill in one install.

1. 1 Add the WaveSpeed marketplace $claude plugin marketplace add WaveSpeedAI/claude-plugins Copy
2. 2 Install the plugin $claude plugin install wavespeed@wavespeed Copy
3. 3 In Claude Code, run /mcp, pick wavespeed and approve the connection in your browser /mcp Copy

Then ask: “Make a 16:9 hero image of a cyberpunk skyline and tell me the price first.”

Other ways

Only the MCP server, without the skill:

$claude mcp add --transport http wavespeed https://api.wavespeed.ai/mcp

GitHub

Three commands and you're shipping

1. 1

   Install

   Node 18+ required. One global install gives you the wavespeed binary everywhere. $npm install -g @wavespeed/cli Copy
2. 2

   Sign in

   Opens the access-key page in your browser. We auto-detect a copied key and validate it live. $wavespeed login Copy
3. 3

   Run any model

   Same shape for all models. Pass the model ID and inputs — get a CDN URL back, or pipe --json to your scripts. $wavespeed run wavespeed-ai/z-image/turbo -p "a cat" Copy

• Teach your coding agent the CLI

  Installs a skill so Codex, Claude Code, Cursor and other agents know how to call it. $wavespeed skill install

• No Node.js?

  On macOS and Linux, install a standalone executable instead. On Windows, download it from GitHub Releases. $curl -fsSL https://raw.githubusercontent.com/WaveSpeedAI/wavespeed-cli/main/install.sh | bash

• Stay up to date

  Upgrade to the latest version with one command. $wavespeed upgrade

• Use WaveSpeed models in your coding agent

  Configure Claude Code, Codex, OpenCode and more to use the WaveSpeed LLM API. Undo any time. $wavespeed setup

Built for humans and agents

Dynamic per-model help

Run wavespeed run <model> -h and the CLI fetches that model's input schema live and prints the exact flags it accepts — no guessing.

Aliases that ship with your repo

Define named shortcuts in wavespeed.json that bundle a model + default inputs. Teams pull the file, type wavespeed run hero, get on-brand output.

Agent-native by design

Every command supports --json on stdout and progress on stderr. Drop a SKILL.md so Codex, Claude Code, Claude Cowork, and Cursor know how to call it.

No magic, no daemon

The CLI never mutates your prompt and never opens a background process. It calls api.wavespeed.ai directly. Open source, MIT licensed.

What a real session looks like

From catalog browse to JSON output — three commands.

# 1. browse the catalog
$ wavespeed models --type text-to-image --popular

# 2. inspect a model's inputs — schema fetched live
$ wavespeed run wavespeed-ai/z-image/turbo -h
  * prompt          string
    size            string   default: "1024*1024"
    strength        number   default: 0.6   range: 0..1
    output_format   string   default: "jpeg"   enum: jpeg | png | webp

# 3. run — JSON out so you can pipe it
$ wavespeed run wavespeed-ai/z-image/turbo \
    -p "a cyberpunk skyline at golden hour" --json
{
  "model": "wavespeed-ai/z-image/turbo",
  "outputs": ["https://example.com/output/skyline.jpeg"],
  "elapsed_ms": 4718
}

Generate images and videos from Claude Code and Codex

Install the CLI and its skill once, then just ask. Your coding agent runs the commands and saves the files in your project.

Claude Code · WaveSpeed CLI

Make a 16:9 hero image for my coffee shop's landing page: cozy, sunrise, room for a headline.

Here's your hero image: warm sunrise light on the storefront, framed wide for a banner. · Cost $0.045

generate image

$ wavespeed generate image "cozy coffee shop at sunrise, warm light in the windows, room for a headline" --aspect-ratio 16:9 --download

Nice. Write "Open at 7" on the front window.

Done: "Open at 7" is hand-lettered on the window, everything else unchanged. · Cost $0.045

edit image

$ wavespeed edit image "add the words \"Open at 7\" in white hand lettering on the front window" --image @./hero.jpg --download

Generating images and videos from your coding agent

How do I generate images from Claude Code?

Install the CLI with npm install -g @wavespeed/cli, sign in with wavespeed login and add the skill with wavespeed skill install. Then ask Claude Code for an image; it runs wavespeed generate image and saves the file in your project.

How do I generate images and videos from Codex?

Install and sign in to the CLI, then run wavespeed skill install so Codex knows the commands. Ask Codex for an image or a video and it runs them for you, or install the WaveSpeed plugin from /plugins.

What does the agent actually run?

Plain commands you can read and repeat: wavespeed generate image, wavespeed edit image, wavespeed generate video. Local files are passed with @./file and the results are downloaded with --download.

Does it work in Cursor and other coding agents?

Yes. Any agent that can run terminal commands can use the CLI, and the skill teaches it the commands. Agents that speak MCP can also connect to the hosted server instead.

Can I generate videos too?

Yes. wavespeed generate video creates a clip from text, a start image or reference images, and wavespeed edit video and extend video change or continue an existing one.

How much does it cost?

You pay per generation from your WaveSpeed balance, with no subscription. Add --estimate to see the price before anything runs; results report what they cost.

What your agent can do over MCP

Shortcut tools for everyday image and video work, plus the full catalog: search models, read a model's real inputs, quote the price, then run it.

generate_image

Create an image from a prompt, optionally guided by reference images

edit_image

Edit, restyle or combine existing images

generate_video

Make a video from a prompt, a first frame or references

edit_video

Change what happens in an existing video

extend_video

Continue a video with new footage

list_models

Search the live catalog by text or type

get_model_schema

A model's real inputs, required fields and defaults

run_model

Run any model; local @file inputs upload automatically

get_price

Quote the cost before you spend

get_balance

Check your credit balance

upload_file

Turn a local file into a hosted URL

get_prediction

Pick up a long-running generation by its id

list_predictions

Browse your run history

list_billings

See itemized charges and refunds

get_usage

Check spend and usage by day and model

delete_predictions

Delete finished runs from your history (asks you first)

upload_file is only in the local server. The remote server can't read your files: pass public URLs, use create_upload from a tool that can send HTTP requests, or, in ChatGPT, attach the file to the chat.

Open source. Ready today.

MIT licensed, scoped under @wavespeed on npm, hosted on GitHub. Issues and PRs welcome.

Prefer a graphical app? Try WaveSpeed Desktop See every way to connect WaveSpeed