What Is MCP in AI? How Aristotto MCP Works

MCP, or Model Context Protocol, is an open standard that lets AI applications connect to external tools and services instead of being limited to what they can do inside the conversation itself.

With Aristotto MCP, compatible AI clients can access Aristotto's creative generation capabilities. You can discuss an idea with your AI assistant, ask it to check available models, prepare an image, video or audio generation, review the credit cost and approve the generation without manually rebuilding the request somewhere else.

The important thing to understand is that MCP is not an Aristotto-specific technology. It is the connection standard. Aristotto MCP is what connects that standard to Aristotto's creative tools.

What does MCP mean in AI?

MCP stands for Model Context Protocol.

The simplest way to understand it is as a common connection between an AI assistant and an external service.

An AI assistant may be very good at helping you develop an idea, write a prompt or plan a campaign. But without access to an external creative tool, it can only tell you what you could create.

MCP changes that.

A compatible service can expose its capabilities through MCP. The AI client can discover those capabilities and use them when they are relevant to your request.

This means the conversation can move from:

Help me come up with a concept for this campaign.

to:

Now create the first visual using Aristotto.

The assistant is no longer only discussing the task. It has a connection to a service capable of carrying it out.

What is an MCP server?

The phrase "MCP server" can sound more technical than the actual experience.

For a creator, think of it as the part that gives your AI client access to another service.

Once Aristotto is connected, the AI client can understand which Aristotto capabilities are available and what information is needed to use them.

You do not need to explain Aristotto's internal systems every time you want to create something.

What does Aristotto MCP do?

Aristotto MCP brings Aristotto's image, video and audio generation capabilities into compatible AI clients.

Imagine you are planning a product campaign in ChatGPT, Claude or another compatible client.

You develop the idea, decide what the first video should look like and then write:

Create a 9:16 product video from this image with a strong opening shot. Check which Aristotto models fit the idea before preparing the generation.

The assistant can check the capabilities currently available through Aristotto rather than simply suggesting a model from memory.

It can then prepare the generation and show you what will be used before the job starts.

That is one of the most useful differences between MCP and a normal conversation with an AI assistant: the conversation can lead directly into an action.

How to connect Aristotto MCP to your AI client

Connecting Aristotto MCP does not require you to create or manage an API key.

You add the Aristotto MCP server to a compatible client and then sign in with your existing Aristotto account.

The exact menus vary between ChatGPT, Claude, Cursor and other compatible MCP clients, but the basic process is the same.

Step 1: copy the Aristotto MCP server URL

Use the Aristotto MCP server URL:

https://aristotto.ai/api/mcp

You only need to add the connection to your client once.

Step 2: add Aristotto to your AI client

Open the MCP, connectors or integrations area inside your AI client.

Add a new remote MCP server or custom connector and paste the Aristotto MCP URL.

Different clients use slightly different terminology. One may call it a connector, another may call it an MCP server or integration.

The underlying connection is the same.

Step 3: connect your Aristotto account

After adding Aristotto, select Connect.

Your browser will open the Aristotto sign-in flow.

Sign in with your existing Aristotto account, review the requested access and approve the connection.

You do not need to copy an API key or manually handle access tokens.

Step 4: start using Aristotto in the conversation

Once Aristotto is connected, you can start with a normal request.

For example:

Check which Aristotto image models are available and recommend one for a realistic fashion campaign.

Or:

I want to create a cinematic 10-second video from this image. Check the available Aristotto models first and recommend the best fit.

Or, if you already know which model you want:

Use Seedance 2.5 through Aristotto to create a 16:9 cinematic video from this image.

The assistant can inspect the capabilities available through the connection and prepare the generation.

Before the generation starts, you can review the settings and credit cost.

Does setup look the same in every AI client?

No.

The basic connection is the same, but the interface depends on the client you use.

ChatGPT, Claude, Cursor and other compatible MCP clients may put MCP connections in different menus and use slightly different names for the same step.

Some clients may also display generation and approval controls differently.

The important part is that they can connect to the same Aristotto MCP server.

We will cover individual client setup in a separate Help Center guide with step-by-step screen recordings.

How does an Aristotto MCP generation work?

Once the connection is active, the process starts with your request.

The AI client can first check the current Aristotto catalog and understand which generation capabilities are available.

It then prepares the generation based on your request.

Before anything starts, Aristotto provides a quote showing the generation settings and credit cost. You approve that quote before the job begins.

Once the generation starts, the client can follow its status and retrieve the result when it is ready.

The result is also a normal Aristotto generation, so it appears in your Aristotto library rather than living in a separate MCP-only system.

What can you create with Aristotto MCP?

Aristotto MCP is not limited to one model or one media type.

The connection can give compatible AI clients access to Aristotto's available image, video and audio generation capabilities.

That means MCP can become useful across a much larger creative process.

You might begin by discussing an advertising concept with your AI assistant. Once the concept is clear, you could create the first image. Then you could use that image as part of a video generation and continue working from the resulting assets.

The important difference is that more of this can happen as part of the same conversational process.

You are not limited to asking the assistant to write prompts for tools you then need to open manually.

Can Aristotto MCP choose a model for you?

A compatible AI client can inspect Aristotto's current generation capabilities before preparing a job.

That means you do not necessarily need to know the model name before you begin.

You can describe what you need:

I need a realistic vertical product video with strong motion and native audio. Check the available Aristotto models and tell me which ones fit before preparing anything.

The assistant can first inspect the available capabilities and use that information when helping you choose.

This is particularly useful because AI model catalogs change quickly.

You can still specify a model directly whenever you already know what you want.

Does the AI automatically spend your credits?

No generation begins simply because the assistant suggested one.

Aristotto prepares the generation first so you can review the settings and credit cost before approving it.

If the model, settings or price are not what you expected, you can change the request before approving.

This is an important part of the workflow because it lets the assistant handle more of the preparation without removing your control over generation costs.

Do you need an API key?

No.

Aristotto MCP uses your Aristotto account connection.

Your compatible AI client opens the Aristotto sign-in flow, you sign in, review the requested access and approve the connection.

For creators, that removes one of the more technical parts of traditional integrations.

You can use an agentic creative workflow without needing to generate, copy and manage a separate API key.

What is the difference between MCP and an API?

An API is a way for software to communicate with another service.

MCP is a standardized way for AI applications to discover and use tools and capabilities exposed by connected services.

For a creator, the practical difference is easier to understand than the technical one.

A traditional API integration usually needs to be built.

With MCP, a compatible AI client already understands the protocol. You connect an MCP server, authorize access and the client can discover the tools that service exposes.

MCP does not replace APIs everywhere. It makes connecting tools to AI applications more standardized.

Is MCP the same as an AI agent?

No.

An AI agent is the system reasoning about the task and deciding which actions may be useful.

MCP is the connection that gives that system access to external tools.

For example, an agent might decide that your campaign needs an image before a video can be created.

It can inspect Aristotto's available image capabilities through MCP, prepare a generation and ask you to approve it.

The agent makes decisions.

MCP gives it a standardized way to interact with Aristotto.

Why would a creator use MCP?

The biggest advantage is not saving one or two clicks.

It is keeping more of the creative process connected.

Imagine you are developing a launch campaign with an AI assistant.

You discuss the audience, concept, hook, visual direction and script. Once you have a concept you like, you ask the same assistant to create the first visual through Aristotto.

You review the result together.

You might then say:

Keep this direction, but make the next version feel less polished and more like a real phone-shot ad.

The conversation already contains the context behind the campaign. You do not have to start again from an empty generation interface and reconstruct everything manually.

That connection between reasoning, context and creation is where MCP becomes much more useful than simply adding another generation button.

Can you use reference images through Aristotto MCP?

Yes, although the exact experience depends on the MCP client.

Compatible clients with media upload support can use reference media in generation workflows. Existing assets from your Aristotto library can also be reused.

This allows workflows to begin with something you already have rather than always starting from a text prompt.

For example, you could upload a product image and ask your assistant to help determine which available video model would be suitable for turning it into an advertising clip.

Which AI clients work with Aristotto MCP?

Aristotto MCP is designed for compatible clients that support remote MCP connections and account authorization.

This includes ChatGPT, Claude, Cursor and other compatible MCP clients.

The exact setup and available interface features can vary between clients.

That means Aristotto MCP may look slightly different depending on where you use it, even though it is accessing the same underlying Aristotto capabilities.

Where do generations created through MCP go?

MCP does not create a separate media library.

Generations created through Aristotto MCP are normal Aristotto generations and appear in your Aristotto library.

This is useful if you start something conversationally and later want to continue working with the result directly inside Aristotto.

MCP is another way into the same creative ecosystem rather than a separate product with disconnected assets.

Is Aristotto MCP replacing the Aristotto platform?

No.

MCP is another way to access Aristotto's creative capabilities.

Working directly in Aristotto makes sense when you want to use areas such as Workspace, Workflows, Video Studio or Marketing Studio.

MCP becomes particularly useful when the creative process starts somewhere else.

You might be planning inside ChatGPT or Claude and want Aristotto generation to become part of that existing process rather than moving the entire project to another interface.

Both approaches can work together.

What should you check before approving a generation?

Check the generation the same way you would if you were preparing it directly inside a creative platform.

Look at the selected model, input media, format, settings and credit cost.

If something does not match what you intended, change it before approving.

The goal of MCP is not to remove you from the creative process.

It is to let the AI handle more of the repetitive preparation while you remain in control of the important creative and spending decisions.