Vixlens

Model Context Protocol

Vixlens for your AI agent

Connect Claude, Cursor or any MCP client to Vixlens, then ask for a video or an image in plain language. Your agent picks the model, runs the generation and brings back the file.

https://api.vixlens.com/mcp
Add to ClaudeGet an API key

Connecting an agent is included on the Studio plan. See pricing.

Just ask

You

“Make a 5-second vertical video of a matcha latte being poured, with Seedance 2.5.”

You

“Generate three square product shots of this sneaker on a concrete floor.”

You

“Animate my logo photo into a short looping clip, then send me the link.”

Setup

Connected in a minute

  1. Add Vixlens to your client

    A custom connector in Claude, or one command or config block in Claude Code, Cursor or VS Code.

  2. Approve access

    Claude sends you to Vixlens to sign in and allow it. Other clients use a key from Settings → API Keys.

  3. Ask for what you want

    Your agent picks a model, queues the generation, waits for it and hands you the result link.

In claude.ai or the Claude app: Settings → Connectors → Add custom connector. Paste this URL, click Add, then Connect and allow access. No API key needed.

https://api.vixlens.com/mcp

For the other clients, replace sk_live_... with a key from Settings → API Keys. Keep it out of shared repositories: anyone holding it can spend your credits.

Tools

What your agent can do

Nine tools, the same ones the web app is built on. Only the three marked below spend credits, at the same price as the web app.

list_models

Every image and video model, with its inputs and settings.

generate_imageUses credits

Queue a text-to-image generation.

generate_videoUses credits

Queue a video from text, or from a start image.

list_presets

The ready-made video recipes from the presets gallery.

generate_from_presetUses credits

Run a preset on your own photo.

get_generation

Status and result link; can wait for it to finish.

list_generations

Your recent generations, newest first.

upload_image

Store an image from a link or file to use as input.

get_account

Your plan and the credits you can still spend.

Queued, not blocking

Generations return a job right away. Your agent follows it with get_generation, which can wait up to 25 seconds per call, and checks again for longer videos.

Same credits, same models

Every model in the web app, billed from the same balance at the same price. Failed generations are refunded automatically.

Your key, your limits

Each key allows 120 requests a minute and can be revoked from Settings at any time. Prefer raw HTTP? Use the REST API.

FAQ

Questions

What is MCP?
The Model Context Protocol is an open standard that lets AI assistants use outside tools. Once Vixlens is connected, your assistant can generate media for you the same way it reads files or searches the web.
Does it cost extra?
No. Generations made through MCP cost the same credits as in the web app, come out of the same balance, and are refunded automatically if they fail.
Which clients work?
Claude as a connector (web, desktop and mobile), plus any client that supports remote MCP servers over HTTP: Claude Code, Cursor, VS Code, Windsurf and others.
Can I add it as a connector in Claude?
Yes. In Settings → Connectors, add a custom connector with the endpoint above, then sign in to Vixlens and allow access. No API key needed, and it works on claude.ai, desktop and mobile.
Where do my results go?
Into your Vixlens gallery, like anything you make in the web app, so you can open, edit or share them there too.
Is Unlimited available through MCP?
No. Unlimited runs are a web-app benefit, so MCP and API generations always use credits.

Start generating from your agent

Add the connector, approve it, and ask.

Set it up