Syllaby MCP Server · Now live
The Syllaby MCP Server
for any AI client
Connect once, then write scripts, render faceless videos and build storyboards from an ordinary conversation in Claude, Claude Code or ChatGPT.
Copy the Syllaby connector URL
You will paste this URL into Claude in the next step
Settings → Connectors
In claude.ai or Claude Desktop open Settings → Connectors → Add custom connector, name it Syllaby and paste the URL
Open Claude ConnectorsConnect, sign in and start
Sign in, then ask Claude to generate a script or a faceless video
Start creatingCopy the Syllaby connector URL
You will paste this URL into ChatGPT in the next step
Settings → Connectors
In chatgpt.com open Settings → Connectors → Add custom connector, name it Syllaby and paste the URL
Open ChatGPT ConnectorsConnect, sign in and start
Sign in, then ask ChatGPT to generate a script or a faceless video
Start creatingCopy the Syllaby connector URL
You will paste this URL into Claude Code in the next step
Settings → Connectors
In Claude Code CLI open Settings → Connectors → Add custom connector, name it Syllaby and paste the URL
Open Claude Code ConnectorsConnect, sign in and start
Sign in, then ask Claude Code to generate a script or a faceless video
Start creatingAs Featured In




















What is the Syllaby MCP Server?
MCP, the Model Context Protocol, is an open standard that lets AI clients call external tools directly. The Syllaby MCP server exposes Syllaby’s video engine as tools any client can use.
You ask
Describe the video you want
Your AI client
Claude, Claude Code and ChatGPT
Syllaby MCP
Authenticates and routes the call
Syllaby engine
Script, visuals and render
Back in chat
Finished video, ready to publish
Works with the AI tools you already use
Add the Syllaby MCP server to any MCP-compatible client, then create videos from an ordinary conversation.
Every tool in the Syllaby MCP Server
27 tools across faceless video, text-to-video, storyboards, assets and credits, callable from any MCP client. Eighteen are free to call; nine charge credits.
Conversational video creation
Ask for a faceless video and the server drafts it, writes the script, renders it and hands back the file.
Script first, render after
Generate a script in chat, refine it in place, then render that same draft without leaving the conversation.
Storyboards, scene by scene
Build a storyboard, edit a single scene, chain one scene into the next, then render the whole thing.
A link becomes a video
Point it at an article and it pulls out the script and the images, then builds the video from them.
Price it before you spend
Ask what a render will cost and check what you have left, before anything is charged.
Every tool in the Syllaby MCP Server
27 tools across faceless video, text-to-video, storyboards, assets and credits, callable from any MCP client. Eighteen are free to call; nine charge credits.
Connect Syllaby to your AI client
The documented ways in. Add a custom connector and authorize over OAuth, or pass a bearer API key for headless environments.
claude.ai
Connector URL
- Settings → Connectors → Add custom connector
- Name it Syllaby and paste the URL
- Authorize via the OAuth prompt
Claude Desktop
Connector URL
- Settings → Connectors → Add custom connector
- Name it Syllaby and paste the URL
- Complete the OAuth sign-in
ChatGPT
Developer mode
- Enable developer mode in Settings → Connectors → Advanced
- Add a custom connector with the server URL
- Complete the OAuth sign-in
Claude Code
OAuth
- Run the command in your terminal
- Approve the OAuth flow in the browser
- The server is live in that session
https://api.syllaby.io/mcp
Claude Code · headless
Bearer token
- Skip OAuth with an API key instead
- Pass it as an Authorization header
- Generate the key in account settings
https://api.syllaby.io/mcp \ --header "Authorization: Bearer slb_..."
Any MCP client
Connector URL
- Add a custom connector or remote MCP server
- Point it at the Syllaby server URL
- Authenticate over OAuth or a bearer key
Built for developers, agencies and content teams
Whether you are wiring Syllaby into an internal assistant, running client work at volume, or you simply want video creation where you already chat, the MCP server fits the way you already work.
Content teams
You plan content in one place and want to create without switching tools.
- Turn an idea or a link into a video in the same thread
- Draft, refine and render a script without leaving chat
- Export the finished video straight from the thread
Agencies
You run video work for several clients and need it repeatable, not bespoke every time.
- Brief in, finished video out, from one chat
- One connector URL for every client account
- Check cost before you commit to a render
Developers
You are building an agent, copilot or internal tool and want video generation inside it.
- One connector URL, no API client to write
- Predictable tool schemas across 27 tools
- Works in Claude, Claude Code and ChatGPT
Popular workflows you can build
In MCP the trigger is the prompt. Type one of these and the server chains the tools for you. No builder, no wiring.
Blog post → faceless short
- 2 min
- “Turn this blog post into a 60-second faceless video.”
scrape_script
create_video
render_video
export_video
- A captioned vertical short, rendered and ready
One sentence → one clip
- 1 min
- “Make me a 10-second clip of a lighthouse in a storm.”
generate_text_to_video
- A single clip, straight from a sentence
Idea → multi-scene storyboard
- 4 min
- “Storyboard a 45-second explainer about Roman aqueducts.”
generate_storyboard
generate_storyboard_media
- Every scene generated, then rendered as one video
Scene → the next scene
- 3 min
- “Continue that scene, carrying the last frame into the next shot.”
chain_storyboard_scene
generate_storyboard_media
- The next shot picks up where the last one ended
One bad shot → replaced
- 3 min
- “Regenerate the third clip and leave everything else alone.”
regenerate_video_asset
render_video
export_video
- One asset swapped, the rest untouched
Cost check before you commit
- 1 min
- “How many credits will four 60-second faceless videos cost me?”
- A price before anything renders
Explore the Syllaby MCP Server
The whole surface in one place: the command you run, what you say once it is connected, and the rules that govern access.
1# add the hosted server2claude mcp add --transport http \3 syllaby https://api.syllaby.io/mcp45# or skip OAuth for CI6claude mcp add --transport http \7 syllaby https://api.syllaby.io/mcp \8 --header "Authorization: Bearer slb_..."
Authentication
OAuth by default, or a bearer API key for headless use
Authorization: Bearer slb_…
Server URL
Every MCP client points at this one endpoint
https://api.syllaby.io/mcp
API keys
Generate a key in your account settings
Manage keys →
Tools
27 callable tools. Eighteen free, nine charge credits.
Browse tools →
Error handling
Consistent error shapes with clear codes
4xx / 5xx
Subscription
An active paid subscription. Trial accounts are rejected.
PAID-SUBSCRIPTION-REQUIRED
Full MCP documentation
Every tool, parameter, error code and best practice in the complete reference.
What it costs to run
Eighteen of the 27 tools are free to call. Nine charge credits, and that list is closed. These are the rules the server actually enforces.
Eighteen free, nine that charge
Reads, estimates, drafts, scene edits and scene chaining cost nothing. The charging list is closed: these nine tools and no others.
Short renders cost the same
Renders bill a minimum of one minute, so a 20-second and a 50-second render on the same engine cost the same.
No tool sells you credits
Nothing on this server tops up a balance or upgrades a plan. If you are short, the agent says so and points you at your account.
Trials can connect, but not call
A trial account authenticates fine, then every tool call is rejected until the subscription becomes paid.
Other ways in
The hosted server is one route. Two more are documented, both installed from npm and both authenticating with the same bearer API keys.
Or skip the chat entirely
The whole API as curated terminal commands, plus a raw escape hatch for every endpoint. Same bearer keys, same rate limits.
Run it on your own machine
The same tools as a local MCP process instead of the hosted server. Installable from npm, the MCP Registry, Smithery, or as a desktop extension.
Three ways to create with Syllaby
Use the app by hand, drive it from the AI client you already talk to, or build straight against the API. Same engine underneath, three different ways in.
Syllaby API
Build Syllaby into your own product.
- REST endpoints with full control
- Webhooks and your own custom logic
- Own the interface end to end
Best for Product and engineering teams
Syllaby MCP
Create from the AI client you already use.
- Works in Claude, Claude Code and ChatGPT
- One connector URL, no code to write
- Chains the tools for you from a prompt
Best for Teams, agencies and power users
Syllaby App
Create by hand in a visual interface.
- Click, preview and edit as you go
- Every feature in one place
- Nothing to set up, just log in
Best for Creators and marketers
MCP questions
What the Syllaby MCP server is, which clients it works with, and what it costs to run.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Yes. Syllaby runs a hosted remote MCP server at https://api.syllaby.io/mcp that exposes 27 tools to any MCP-compatible client.
Still have questions?
Contact our support team