Shotstack MCP Server & CLI

A free, open source Shotstack MCP server and CLI that lets Claude Code, Codex, Cursor and other AI agents render videos from templates, make AI clips and voice-overs, and manage hosted files.

Navid Moazzezby Navid Moazzez·Updated Oct 5, 2026·14 min read·
Your computer
  1. 1

    Add it with a single command in a terminal.

    Put your own API key in place of the example, and set the environment to match it.

    claude mcp add shotstack \
    -e SHOTSTACK_API_KEY=YOUR_API_KEY \
    -e SHOTSTACK_ENV=stage \
    -- npx -y @thenavidm/shotstack-mcp-cli@latest
  2. 2

    In Claude Code, type /mcp. Shotstack should show as connected.

    /mcp

Needs Node 22 or newer and a Shotstack API key. New accounts get 10 free credits. Get your key in a minute.

It's free and open source. If it saves you time, star it on GitHub, so more people find it.

Rate this tool

This free Shotstack MCP server and CLI gives your AI real access to your Shotstack account: renders, templates, AI generation, hosted files and source files. It renders videos from an edit or a template, makes images, clips and voice-overs with AI models, checks any media file, and manages the files Shotstack hosts for you.

It's one install with two ways in. Claude, Codex, Cursor or any other MCP app calls its 23 tools for you, and the same tools work as a CLI that agents like Claude Code, Codex and OpenCode run, or that you type yourself.

In this guide, I'll show you:

  • What the Shotstack MCP server & CLI is, and what you can ask it
  • How to install it and check that it works
  • When to use the MCP server, and when the CLI
  • Every tool, and how to keep your account safe

Every tool on this page comes from the server itself, and every cost was measured in Claude Code.

key_takeaways.mdTL;DR

Key takeaways

It gives Claude, Codex, Cursor and any other MCP app 23 tools for Shotstack: render videos from an edit or a template, make AI images, clips and voice-overs, and manage the files Shotstack hosts.
A new Shotstack account gets 10 free credits, and sandbox renders are free with a watermark, so you can try it before you pay.
Every render, AI generation, upload and delete waits for your approval.
The same tools run as a CLI, which costs your AI nothing until it runs.

What is the Shotstack MCP server & CLI?

The Shotstack MCP server & CLI is a free, open source program that lets AI agents make and manage videos in Shotstack for you, in two ways. The MCP server is what an AI app like Claude, Codex or Cursor connects to, through MCP (Model Context Protocol), the open standard AI apps use to call outside tools.

You ask in plain language. Your AI picks the right tool, and the server makes the call to Shotstack's own API.

The CLI is the same program as commands. shotstack-cli list-templates runs the same code your AI runs when you ask which templates you have, whether an agent like Claude Code runs it or you do.

In an AI OS, the layer that connects your AI to your context, your data and your tools, a server like this is how your AI reaches Shotstack. I explain how the parts fit together in how to build an AI operating system, and here's how I use AI myself.

What can you ask it?

Once it's set up, you ask the way you'd ask an assistant. These are real prompts it handles:

Try asking
Which AI models can I generate with, and what does each one cost?
Render my promo template with the title "Spring sale", in the sandbox first.
Is my last render done? Send me the link.
How long is the video at this link, and what size is it?
Make a 5-second clip of waves at sunset with the cheapest video model.
Delete the thumbnail and poster from yesterday's render.

The second one shows why this is useful. Your AI fills in the template and renders it in the sandbox, where renders are free, so you see it before a production render spends a credit.

It works well next to my Creatomate MCP server and Wistia MCP server, so your AI can render videos from your Creatomate templates and upload the finished videos to Wistia in the same chat. Every server I've made is on my MCP servers page.

For the videos themselves, my free video resizer and video caption generator help too.

How to install the Shotstack MCP server

Pick your app in the box at the top of this page. Each one is a single command or a pasted block, and they all run through npx, so there's nothing to install first.

Windsurf, Zed and Cline take the same server, though they don't have a tab in the box yet. Follow the box's Cursor tab, and put its block here instead:

AppWhere the block goes
WindsurfIn the Cascade panel, open the Actions menu (…) and choose Open MCP config file, then paste the block.
ClineClick the MCP Servers icon in the Cline panel, open the Configure tab, click Configure MCP Servers, and paste the block.
ZedGo to Settings, then AI, then MCP Servers, and choose Add Server, then Add Local Server. Enter the block's command, arguments and variables, and Zed saves them under context_servers.
Before you start0/3

Watch out: Shotstack gives you two keys, one for the sandbox and one for production. The server uses production unless you set SHOTSTACK_ENV=stage, so the sandbox key needs that setting too.

Set up your Shotstack account

It takes a minute.

Get your API key0/4

Sandbox renders are free and carry a watermark. You still need at least one credit in your balance to use the sandbox, and AI generation spends credits there too. When you're ready for clean renders, switch to the production key and set SHOTSTACK_ENV to v1, or leave it out.

A credit is one minute of rendered video, and a new account gets 10 of them, valid for 30 days.

To keep the key out of your app's config, save it alone in a file only you can read, and point SHOTSTACK_TOKEN_FILE at that file's full path instead.

Check that it works

Run the doctor with --network. It checks your settings, then reads Shotstack's model list to prove the key works, without rendering anything.

npx -y @thenavidm/shotstack-mcp-cli doctor --network

When every line has a check mark, restart your app and ask it which templates you have. A failed key check usually means the key and SHOTSTACK_ENV don't match: the sandbox key needs stage, and the production key needs v1.

In Claude Code, /mcp also shows whether the Shotstack server is connected. It's one of the Claude Code commands worth knowing.

Use the Shotstack CLI

The CLI is the same 23 tools as commands. AI agents that run commands, like Claude Code, Codex and OpenCode, use it on their own, and you can type the same commands in a terminal or a script.

Every tool name becomes a command with dashes, so render_template runs as shotstack-cli render-template.

npm install -g @thenavidm/shotstack-mcp-cli
shotstack-cli
shotstack-cli list-models
shotstack-cli list-templates
shotstack-cli render-template --id YOUR_TEMPLATE_ID --merge '{"find":"TITLE","replace":"Spring sale"}' --confirm
shotstack-cli get-render --id YOUR_RENDER_ID

The bare shotstack-cli lists every command, shotstack-cli which <words> finds the command for a task, and shotstack-cli <command> --help shows what a command takes. Renders, AI generation, uploads, transfers and deletes need --confirm, the terminal's version of your approval.

shotstack-cli install <app> adds the server to Claude Code, Codex, Claude Desktop, Cursor, VS Code or Gemini CLI, in each app's own format.

These are the flags you'll use most:

FlagWhat it does
--jsonPrints JSON
--compactPrints the JSON on one line
--select a,b.cKeeps only those fields, and a dotted path goes deeper
--agentMachine mode: compact JSON, no prompts, and it never confirms anything
--confirmLets a write that asks first go ahead
--dry-runShows what a write would do, without doing it

A script can branch on the exit code:

Exit codeWhat it means
0It worked
1Something unexpected went wrong
2The command was typed wrong, or a write needed --confirm
3It wasn't found
4Shotstack rejected the credentials
5Shotstack failed or couldn't be reached
7You hit a rate limit, so wait and try again
10Nothing is set up yet

MCP server or CLI: which one?

Both are the same program with the same 23 tools. The difference is when your AI pays for them, in the tokens it reads.

Every message, in an app that loads every tool
MCP server
154,100 tokens
CLI
Nothing
Every message, in Claude Code
MCP server
480 tokens
CLI
Nothing
When Shotstack comes up
MCP server
Nothing more, or in Claude Code the tools it picks
CLI
1,300 tokens, once
20 messages with Shotstack in one, every tool loaded
MCP server
3,082,000 tokens
CLI
1,300 tokens
Works in Claude Desktop's chat
MCP server
Yes
CLI
No, there's no terminal there
Works in a script, a cron job or CI
MCP server
No
CLI
Yes

An app that loads every tool up front sends the full list with every message, whether you mention Shotstack or not. Claude Code doesn't by default: its tool search sends only the tool names and the server's instructions, and loads a tool's full definition when your AI reaches for it.

The CLI costs nothing until Shotstack comes up. Then your agent reads its skill file once and runs the commands it needs. With the skill added, Claude Code also lists its one-line description with every message, about 60 tokens.

When the whole conversation is about Shotstack, the gap closes. Then the MCP server is the better experience, because you ask in plain language and your AI never has to look up a command.

How to spend less

In Claude Code, leave tool search on, which it is unless you've set ENABLE_TOOL_SEARCH=false.

Turn the server off when you aren't making videos. In Claude Code that's the /mcp panel, and every AI app has its own switch. SHOTSTACK_READ_ONLY=1 cuts it to the 12 reading tools.

Or install the CLI and connect the server later, on the days it earns its place. Every tool stays in reach, and the standing cost drops to nothing.

Every number here was measured on October 5, 2026 in Claude Code 2.1.286 with Claude Opus 5.5. It's the same one-line prompt with and without the server connected, once with tool search off so every tool loads, and once with Claude Code's default. The skill was measured the same way, and other apps and models count tokens a little differently.

Render, check, then keep

A render is a job. render or render_template hands Shotstack the edit and gets back an ID, and get_render shows where the job is until it's done or failed. Your AI checks that same ID again, so a slow render never turns into a second one.

AI generation works the same way. list_models shows each model and what it costs in credits, generate_asset starts the job, and get_generated_asset follows it. Give the job an idempotency key, and Shotstack won't make it twice within 24 hours, even if it's sent again.

Shotstack hosts what it makes. get_asset_by_render_id lists every file a render made, like the video, its thumbnail and its poster, and each one is deleted on its own.

To bring your own media, ingest_source has Shotstack fetch a file from a link and store it. create_upload_url_file gets a private upload link instead, and since that link works like a temporary password, the server saves it to a new private file rather than showing it to your AI.

Watch out: Shotstack counts each key's requests in one-minute windows, and renders have the tightest limit. After a rate limit error, wait for the minute to pass, and check a render before you send it again, because the first one may already be running.

Every Shotstack tool

The tables below list all 23 tools, grouped by what they're for. Of these, 12 only read, and 11 change something in your Shotstack account or spend credits.

Renders

render
What it does
Renders an edit, a timeline of clips, text and audio, into a video, image or audio file.
Kind
Asks first
get_render
What it does
Shows a render's status, and a link to its file once it's done.
Kind
Reads
probe_media
What it does
Reads a media file's details, like its length, size and codecs, from its link.
Kind
Reads

Templates

create_template
What it does
Saves an edit as a template you can render again.
Kind
Asks first
list_templates
What it does
Lists your saved templates.
Kind
Reads
get_template
What it does
Reads one template in full.
Kind
Reads
update_template
What it does
Changes a saved template.
Kind
Asks first
delete_template
What it does
Deletes a saved template.
Kind
Asks first
render_template
What it does
Renders a template, with its merge fields filled in.
Kind
Asks first

AI generation

generate_asset
What it does
Makes one image, video clip or voice-over from a text prompt with an AI model, which spends credits.
Kind
Asks first
get_generated_asset
What it does
Shows an AI generation's status, and its file once it's done.
Kind
Reads
list_models
What it does
Lists the AI models you can generate with, and what each one costs in credits.
Kind
Reads
get_model
What it does
Reads one AI model, with the options it takes and its cost.
Kind
Reads

Hosted files

get_asset
What it does
Reads one hosted file, like a video, thumbnail or poster.
Kind
Reads
delete_asset
What it does
Deletes one hosted file.
Kind
Asks first
get_asset_by_render_id
What it does
Lists every hosted file a render made.
Kind
Reads
transfer_asset
What it does
Copies a file from a public link to one of your hosting destinations.
Kind
Asks first

Source files and uploads

ingest_source
What it does
Has Shotstack fetch and store a video, image, audio file or font from a link.
Kind
Asks first
list_sources
What it does
Lists the source files Shotstack has stored for you.
Kind
Reads
get_source
What it does
Shows one stored source file and its status.
Kind
Reads
delete_source
What it does
Deletes one stored source file.
Kind
Asks first
create_upload_url_file
What it does
Gets a private upload link, and saves it to a new private file on your computer.
Kind
Asks first

Your accounts

list_accounts
What it does
Lists the accounts you've set up, without their keys.
Kind
Reads

Is the Shotstack MCP server safe?

Renders and AI generation spend credits, and a deleted file or template is gone. So 11 tools ask first: every render, every AI generation, every upload or transfer, and every save or delete. They're the ones marked "Asks first" in the tool tables above.

Reading a render's status, a template or a model never asks, because it changes nothing.

When your AI calls one of them, the call needs a yes first. Where you use it decides who gives it:

  • Claude Code (2.1.246 and later) asks you before every call, in every one of its permission modes.
  • An app that can show forms asks you in a form.
  • Any other app takes your AI's confirm: true as the yes, so tell your AI to check with you first.
  • An agent with no person to ask needs SHOTSTACK_CONFIRM=model, so its confirm: true counts in every app.
  • In a terminal, you add --confirm to the command.

If the answer is no, nothing reaches Shotstack.

Make it read-only

Set SHOTSTACK_READ_ONLY=1 and every tool that asks first disappears from the tool list, leaving 12 reading tools.

Keep a log of every write

Set SHOTSTACK_AUDIT_LOG to a file path. The server writes one line per attempted write, allowed or blocked, with what it was about to do and who approved it.

Watch out: Template text and media details can come from other people, and one can try to give your AI orders. Treat them as data, and use SHOTSTACK_READ_ONLY=1 for an agent working on its own.

Your data

Your requests go straight to Shotstack, with no server of ours in between and no tracking. Your key stays in your app's config, your shell or your token file.

The media links in an edit go to Shotstack, which fetches them to render. The server itself never downloads a render or uploads a file from your computer.

To disconnect, remove the server from your app's config. The keys themselves are on Shotstack's API Keys page.

Connect more than one Shotstack account

Set SHOTSTACK_ACCOUNTS to a JSON list with a name, a key or a token file, and an environment, stage or v1, for each one. SHOTSTACK_DEFAULT_ACCOUNT picks the one that acts when a tool names none, and every tool takes an optional account.

That's also how to keep the sandbox and production side by side: one account called sandbox with the stage key, and one called production with the v1 key.

Shotstack MCP server settings

Every setting is an environment variable, in your app's config or your shell.

SHOTSTACK_API_KEY
Default
none
What it does
Holds your API key
SHOTSTACK_TOKEN_FILE
Default
none
What it does
Points at a private file holding the key, instead
SHOTSTACK_ENV
Default
v1
What it does
stage uses the sandbox, and v1 is production
SHOTSTACK_ACCOUNTS
Default
none
What it does
Lists several accounts, each with its environment, as JSON
SHOTSTACK_DEFAULT_ACCOUNT
Default
the first one
What it does
Picks the account that acts when a tool names none
SHOTSTACK_REQUEST_TIMEOUT_MS
Default
30000
What it does
Gives up on a Shotstack request after this many milliseconds
SHOTSTACK_MAX_RETRIES
Default
2
What it does
Retries a rate-limited read this many times, and never a write
SHOTSTACK_MIN_REQUEST_INTERVAL_MS
Default
150
What it does
Waits this many milliseconds between requests
SHOTSTACK_READ_ONLY
Default
off
What it does
1 hides every tool that changes something or spends credits
SHOTSTACK_ALLOW_DESTRUCTIVE
Default
on
What it does
0 refuses the tools that ask first
SHOTSTACK_AUDIT_LOG
Default
none
What it does
Names a file that gets one line per attempted write
SHOTSTACK_SURFACE
Default
full
What it does
search lists three tools that find, describe and run the rest
SHOTSTACK_TOOL_TIMEOUT_MS
Default
none
What it does
Gives up on any tool after this many milliseconds
SHOTSTACK_CONFIRM
Default
human
What it does
model lets your AI's confirm: true approve on its own, for an agent with no person to ask
SHOTSTACK_HTTP_PORT
Default
8787
What it does
Sets the port for --http
SHOTSTACK_HTTP_HOST
Default
127.0.0.1
What it does
Sets the address for --http, and any other needs a token
SHOTSTACK_HTTP_TOKEN
Default
none
What it does
Sets the bearer token every --http request needs
SHOTSTACK_HTTP_ALLOWED_ORIGINS
Default
none
What it does
Lists other websites allowed to call --http, comma separated
SHOTSTACK_DEBUG
Default
off
What it does
1 prints debug lines

Troubleshooting

Run the doctor first. It names the step that failed and the fix.

What you seeWhat to do
The key check failsThe key and SHOTSTACK_ENV don't match. The sandbox key needs stage, and the production key needs v1
Renders have a watermarkThat's the sandbox. Use the production key with SHOTSTACK_ENV=v1 for clean renders
The sandbox won't renderYou need at least one credit in your balance, even for free sandbox renders
A 429 errorShotstack counts requests per minute. Wait for the minute to pass, and check the render before you send it again
A template renders with the wrong textEach merge field's find value has to match the placeholder in the template exactly
The upload link file is refusedIt never overwrites a file. Pick a new file name, in a folder only you can read
"…so it will not run without confirm: true"That's the safety check working. Approve it when your app asks

If the server doesn't show up in your app at all, run the command your app runs, in a terminal, and read the error.

More free tools for video

Make videos from templates in Creatomate, or host and track them in Wistia.

FAQs about the Shotstack MCP server & CLI

Here are the questions people ask most about the Shotstack MCP server and CLI.

The Shotstack MCP server & CLI is a free, open source program that lets an AI app like Claude, Codex or Cursor make and manage videos in Shotstack.

It has 23 tools for renders, templates, AI generation, hosted files and source files, and the same tools run as a CLI that agents like Claude Code and Codex use, or that you type yourself.

An MCP server is a standard way to give an AI app real access to a tool, so it can act instead of guessing.

MCP stands for Model Context Protocol, and Claude, Codex, Cursor and many other AI apps speak it.

Every server I've made is on my MCP servers page.

The Shotstack CLI is the same program as the Shotstack MCP server, run as commands.

AI agents like Claude Code, Codex and OpenCode run them for you, and you can type them in a terminal or a script.

One npm package installs both.

Use the Shotstack MCP server in an AI app with no terminal, like Claude Desktop's chat, and the CLI in a terminal, a script or a cron job.

In an AI app that can run commands, like Claude Code, Codex or Cursor, the CLI is the lighter choice, because it costs nothing until it runs.

Yes, the Shotstack MCP server is free and open source under the AGPL-3.0-or-later license.

Production renders and AI generation spend Shotstack credits from your own plan, and a new account starts with 10 free credits.

The sandbox makes free renders with a watermark, and production makes clean renders that spend credits.

Each has its own key, and SHOTSTACK_ENV picks which one the server uses.

A Shotstack credit is one minute of rendered video, and a new account gets 10 free credits for 30 days.

AI generation costs depend on the model, and list_models shows each one's price.

Yes, generate_asset makes one image, video clip or voice-over from a text prompt, with any model list_models shows.

It spends credits, even in the sandbox, so it asks first.

Yes, set up two accounts in SHOTSTACK_ACCOUNTS, one with the stage key and one with the v1 key.

Every tool takes an optional account, so you say which one acts.

The Shotstack MCP server works with any app that speaks MCP.

That includes Claude Code, Claude Desktop, Codex, Cursor, VS Code, Windsurf, Zed, Cline, GitHub Copilot CLI, Gemini CLI, OpenCode, OpenClaw, Antigravity and Hermes.

I did.

I'm Navid Moazzez, and I made the Shotstack MCP server & CLI as one of my free tools on navid.me.

You can read more about me.

The code is on GitHub under the AGPL-3.0-or-later license, and the package is on npm.

More MCP servers & CLIs

Related free tools

Navid Moazzez

AI business strategist

Navid Moazzez helps creators and founders master AI and build their own AI Operating System (AI OS) to automate their business and life.

Navid.me is reader-supported. When you buy through links on this site, I may earn an affiliate commission. Read the affiliate disclaimer.

Free AI newsletter

The most actionable AI newsletter for founders

Every week, get proven AI strategies, curated tools, and step-by-step systems to grow your audience, create better content, and build a profitable creator business.

No fluff, no filler, no BS. Just five minutes each week that might level up your online business and life.

P.S. Sign up now to get free access to my ultimate AI tools guide for creators.

Loved by 10,000+ readers
Loved by 10,000+ readers