The new Kling AI MCP plugs one of the best AI video models in the world directly into Claude, Cursor, and Codex. In this video, I show you the two-minute setup, then use it to generate six product showcase videos from a single photo, build a reusable Claude skill for product videos, and run a complete ad creative pipeline: pulling a winning ad from the Meta Ad Library, breaking down why it works, scripting a new version for my own product, and generating the finished 30-second ad without ever leaving the chat. If you sell products, run paid traffic, or do agency work, this is your ad creative pipeline. Try the Kling MCP for free: https://kling.ai/app/mcp/guide
CHAPTERS: 0:00 AI Ad Creatives Made Inside Claude 0:22 What is the Kling AI MCP? 0:50 In This Video 1:04 What is MCP? (Model Context Protocol Explained) 1:28 Why the Kling MCP Changes AI Video Generation 2:30 Who the Kling MCP is For (Ecommerce, Ads, Agencies) 3:15 Follow Along + Get Started Link 3:43 How to Set Up the Kling AI MCP in Claude (2-Minute Setup) 4:38 Testing the Kling MCP Connection 5:05 Demo 1: Generate 6 Product Videos From One Photo 6:24 Building a Claude Skill for High-Converting Product Videos 6:54 Reviewing the AI Product Showcase Videos 7:58 Demo 2: Full AI Ad Workflow (Meta Ad Library Method) 8:41 Analyzing a Winning Video Ad With Claude 8:58 AI Ad Script + Character Design 9:17 Generating a 30-Second Ad With the Kling MCP 10:05 Reviewing the Final AI-Generated Ad 10:52 Kling MCP Pricing, Free Usage & Practical Tips 11:46 Who Should Use This + Final Thoughts
------------------- Clips channel @jonlawclips
I do NOT sell any products of my own. Watch out for scams.
This video demonstrates how to integrate Kling AI into AI agents like Claude, Cursor, and Plod using the Model Context Protocol (MCP). This integration allows users to generate high-quality product videos and full ad campaigns directly within a chat interface, eliminating the need to switch between different platforms and significantly reducing production costs and time.
So everything that you just watched is a sellable ad creative.
An agency will quote you a few thousand dollars for a package like that.
A freelancer would take a week.
And I just made it inside Plod with one message while I was getting my morning coffee.
Now this is the new Kling AI MCP and it plugs Kling,
which is one of the top AI models in the world,
directly into the agents that you and I already use.
So your cursor, Plod, and Codex.
And that turns your agent into something you can actually build a business on
and integrate into the wider staffing of AI that you use.
You're talking product videos at scale, ad variations at will,
and entire campaigns of whatever sort you can imagine from a conversation inside Plod or GPT.
And so in this video, I'll show you the two-minute setup,
how to batch a month of ad creatives from one pop,
and how to run a complete campaign pipeline without ever leaving your chat.
That's going to be the video.
It's going to be a fun one.
And quick context here, if you're not really deep into the AI world.
So MCP stands for Model Context Protocol.
It's kind of like the standard way to plug tools into the AI agent.
So when you connect a tool over MCP, the agent or the AI can actually use the tool.
It's kind of like making it readable to the AI.
So it's not going to describe it.
It's not going to link you to it.
The AI can actually do the work that you want directly via this MCP connection.
So Cling Longchain, their official MCP means,
now they are now a capability inside all of these AI tools from Chatbot.
You describe the video that you want in plain language.
Your agent calls Cling, and the finished videos will come back to you.
That's text to video, image to video.
And this flips the whole model of how we traditionally use these AI tools.
Because until now, every video platform was a destination.
If I wanted to generate a video, if I had an idea, I would go to the website.
I would go find the tool within the platform.
I would type in a prompt.
I would get it back.
Then I would go back to what I was doing,
maybe in the other Chatbot tool in the first place.
And now that's flipped, whereas you can just do that directly in the Chatbot
in what is becoming the nexus or the home of the scaffolding
for all these different AI-enabled workers and then business as well.
Again, your prod, your product, your cursor.
And so that unlocks a whole new set of just capabilities,
things that were not possible before, which is what we're going to cover in this demo.
Now, in terms of who this is for, if you sell products online,
this right here is your ad creative pipeline.
If you run paid traffic, this is creative testing at massive volume.
If you do freelancer agency work, this is now a service you can deliver in an afternoon.
And if you already live in Cursor or Prod or Codec,
this is one more capability that those agents have that you can just tap on when you need it.
The comments write us that production used to be an extremely expensive part.
And when I say used to be, I mean like five years ago,
three years ago even in terms of the quality of these videos.
It's not less time than that.
Production used to be an extremely expensive, time-intensive,
and talent-bounded thing.
And now you can do it basically for free at scale.
With that said, let me just show you.
We're going to jump into the demo in a second.
And before we do that, this video thankfully is sponsored by Kling,
so I can show you guys the empty tool.
And everything I'm about to show you is me actually using it.
It's all real.
Nothing is cut.
And what's best for you to learn from this video is to just follow along and build with me.
Go through the two-minute setup like you're about to see.
So the link to start is in the description or on the screen here.
You can click on that and follow along while we go through this setup that we're about to do.
And so speaking of, let's jump into my computer.
Here we are inside Cloud.
We're going to start with that setup.
It's a very similar process in other platforms.
But within Cloud, let's click on our profile photo, then settings, then connectors.
And then inside of connectors, we'll click add, and then add custom connector.
Now let's go ahead and call this Kling AI.
And then we'll paste this link.
Kling.ai forward slash mcp.
So when you're doing this, it does have to be verbatim these two text blocks right here.
And with that, we'll click add.
And then we'll just go ahead and click connect.
This will connect us to our Kling account.
So I'll continue with Google.
And then right here, we will authorize the connection.
Let's go back to Cloud.
And so now right here, we can see that Kling AI is connected as a custom MCP.
And so there you go.
It's no API keys, no custom code.
It takes 60 seconds on any of these AI platforms, and then you're good to go.
And as an experiment here, let's say, hey, Cloud, what can you do with Kling now?
And we'll make sure to see that connection went through.
Yep.
So we can see Kling AI is connected.
We can generate images.
We can generate videos.
Cloud can upload files straight directly into Kling, and we can manage our account.
And so here's where it gets really fun and into the genuinely impressive use cases here.
So let's go ahead and search for a water bottle, just a basic product photo.
This one right here works.
I'll go ahead and save that.
I'll plug it into Cloud, and then I'll say this.
Take this product photo and use Kling to generate six showcase videos, not photos.
This will be videos.
And this is six videos.
This is something you would need to spend thousands of dollars or at least a few hours
of your time if you can do it in a studio to make.
And we're asking for all six of them at once that could be used for social media, for ads,
for your website.
Let's see if Kling can pull it off.
We can see it found the tools, and then now it's going ahead and calling that Kling NCP.
Now it's uploaded that photo successfully.
Okay, it's completed one of these six.
It is waiting to not overwhelmed API.
Okay, it's submitted all of them, and it's waiting for them to get finished inside of
Kling before it queries the results back into Cloud.
Now you don't have to, but if we jump into the actual Kling side, we can also see those
videos here.
So you can see the literal result of the prompt on the Cloud side in your Kling dashboard.
Now, a quick note while this is wrapping up, the way you're going to get the most use out
of tools and integrations like this is by adding on skills.
Skills are kind of little training sets or instructions for AIs to follow so that they
can do their job better.
So what you can do, and this is very easy, is say something like this to the chatbot,
build me a skill for generating high converting product videos with Kling.
Let's go ahead and get that running, and then we'll install the skills so it can be
used for future generations.
And this is what I suggest you do as well, again, to get the best possible use out of
this tool and anything like it.
Here we can see that's completed.
Now what I can do is just click save skill.
And so this is the Kling product video skill.
If I click manage, I can see it right there.
So again, I recommend you do that.
And what it looks like is just a set of instructions for how to use this better.
So jumping back here, we can see all of those six videos are completed, and we can go ahead
and see the direct links to those as well if we want to.
Here's one of those.
We can see it slowly turning around.
Looks awesome.
They are good and cheap.
And here's another one of those videos.
This is a pen down.
Let's go ahead and see dark background.
I like the look of that.
And then the hand picking it up off the desk.
Okay.
And that looks awesome as well.
So there's our six videos.
And it also tells us very nicely how many credits we have left.
So that in mind, I want to jump into what I would consider a kind of full workflow here
for video generation, in this case, a product ad.
So I'm in the meta ad library here.
I searched for headphones, but I want to find an ad, a video ad that's done well that we
can recreate.
So let's look for something like this.
Basically just putting this headphone sleep mask on.
I think that looks good.
I'm going to click on this.
That gets us over to this page.
And then let's go ahead and download this video.
I'm copying the link going right here.
I'm going to download that video.
Now we have that video downloaded.
Let's plug it back into Quad.
And so this is an example, a ad that performs well based on something you've seen that,
you know, performs well because it's been running on the Facebook ad library for a
minute.
Now I'm going to give it this prompt.
Quad breakdown.
Why?
Let me change it to this ad works.
So pacing, shot structure, hook tone, etc.
Let's go ahead and see what it says there.
Okay.
It's given us the full breakdown.
This is very thorough breakdown.
I like it.
Now what we're going to say is based on that analysis, write an ad script for our blue
water bottle and do character design.
So a model for our ad.
Let's let it do that.
Okay.
Awesome.
So it looks like we have the full script.
We have the consistency.
And now let's go ahead and just make the ad.
So now feed this to cling to generate a 30 seconds for me following the character design
and storyboarding scripting.
So we've gone from a reference, something we know performs well to fleshing that out,
converting that into something that can perform well for us and our product.
And then now we're going to go ahead and use the cling MCP to generate it.
And we can see cloud is taking initiative here.
So it's generating 15 second clips, and then it's going ahead and submitting those photos
for reference on its own.
We can also see the character reference design that it made.
So this is the model that is refeeding into cling to use in those videos to be the model
for our water bottle.
Okay.
We can see that the two halves of it are done.
Let's go ahead and click open link.
So here is part one.
Let's give that a watch.
Looking really good.
Looking human, looking realistic.
We have a different scene.
Some that's in line with the script.
I mean, again, the realism here is really, really good.
That's a beautiful shot with the ice and the water right there.
And so overall, there you go.
That's your ad.
And that was generated completely within clouds.
Whether you want bulk text to image, whether you want text to video, you can do that natively.
You just click on these links and that's how you get your final results.
Now, a few practical notes here so you get the most out of these tools.
First, if you want to produce actual volume, you will need to subscribe to cling.
That's because subscribers can run multiple tasks at once.
So you can bulk.
You can use it for free within these tools.
But if you subscribe and you can run multiple tasks at once.
So for the bulk generation, like we looked at, that's what's going to make your batch
workflows really, really fast.
Again, you can do it completely for free.
You're just limited to that one generation at a time.
Now, MCP generation is priced exactly the same as the platform itself.
And any promotion deals in the app will apply to the MCP tool as well.
And then finally, generated file links expire after 24 hours.
So you should ask your agent to save output as they come in for safety.
They can do that very easily.
And that saves you the annoyance of generating a bunch of stuff, storing the links somewhere.
And then you come back in a week and you're like, Hey, where am I?
So overall, if you sell products, you run ads, you do client work, or you're just a
very happy cursor Codex user who wants videos on tap.
This is 100%, not even a question worth the two-minute setup.
The link to start that is in the description or on the screen here.
You can get started and generate completely for free via MCP, and you can subscribe to
get those higher batch volumes.
And so go ahead and drop a comment with the use case that you are most excited about,
or just one of the new capabilities that's been unlocked with these MCP video connections.
And maybe we'll test that workflow in a future video.
Either way, I appreciate y'all being here and I will see you in the next one.