Meta just released its most capable AI model, which is MuseSpark 1.1.
And according to Zugg, this is a model that can act, it can watch a video, use tools and get it as done for you, not just answer questions.
And later in the video, I'm going to talk about all of that.
Mark Zuckerberg, actually, the way he announced it on the X platform, I think he came back on it after a few years.
And look at what he leads with, a strong agentic and coding model at a very low price through our new Meta model API.
A low price, a paid API.
Now think about this, from a company that for years made free and open source AI its entire identity.
So before I get into what Spark can actually do, I think it'll be interesting to understand what pushed Meta to change the strategy like this.
So let's go back to history and rewind, right?
Because the backstory explains everything.
For years, Meta was the champion of open source AI.
Back in July 2024, Zuck published an essay titled, literally, Open Source AI is the Path Forward.
Free downloadable models were Meta's whole identity, and he positioned the company as the open alternative to closed labs like OpenAI.
Then it changed fast.
Meta brought in Scale AI's Alexander Wang as their chief AI officer, a $14 billion deal.
They stood up a brand new super intelligence lab, and back in April, shipped the first Mewk Spark model, which was closed, no waits, no download available.
That was a real break from open source.
For the first time, right?
And Meta's frontier model were locked behind its own doors, running inside its apps instead of out in the open.
And this week, Meta took the next step, a big upgrade, Mews Spark 1.1.
And for the first time ever, a public paid API with pricing, Zuckerberg called very aggressive.
So Meta didn't just close its best model, it put a price tag on it and open it up for anyone to build it.
Meta didn't just undercut everyone on price.
It's about a quarter of what GPT and Claude cost.
They also made Mews Spark drop straight into the tools developers already use.
OpenAI setup, sure, but also Claude code, Anthropic's own coding tool.
You pointed at Meta's model change basically one line, and now you're suddenly running on Mews Spark at a quarter of the price.
Meta built a side door out of OpenAI and Anthropic and hung a discount sign on it.
This isn't just a new model, it's a direct play for their rivals' customers.
Now, two things to keep honest.
One, Lama isn't dead, it's still open, still downloadable.
What changes, Meta's best models are closed now.
And two, as wild as the flip feels, it makes sense, right?
Meta's spending north of $100 billion a year on AI.
You don't spend that and give your best model away.
And if you want a personal agent in the hands of 3 billion people, you want to own the model doing the work.
So what did they actually build, and is it worth your time?
I started using it for the last couple of days, and I also built a little app on Meta's model.
I handed it a video of anything that you probably want to sell, and asked it to give me a full price, ready-to-post listing on Facebook's marketplace.
So that's one demo that we will see.
Second, I also ran it on Claude code's Anthropic, so that'll be an interesting one.
And third, I wanted to test the real multimodality, so I gave it a picture of the ingredients in the fridge, and it was able to pinpoint the price tag on it, right?
So I'll show you all of the fun parts as we go through, but then, interesting part, as I mentioned earlier, is Meta's shift to the paid strategy versus just keeping it open source.
All right?
Before we dive deep into it, a quick disclaimer, all opinions are my own and do not belong to my employer.
All right?
With that, let's get into it.
All right, so one of the fastest ways to experience Spark is to actually come to Meta.ai and change the mode from Instant to Thinking mode.
So if you ask any question, like, for example, here I'm asking which particular Meta model are you using?
You would see that it is going to confirm that it is using Spark 1.1.
So you just need to change the mode to Thinking, and you will get to leverage the Spark 1.1 model, right?
So it was launched literally a couple of days back, okay?
But then, from a developer perspective, if you want to use it, you need to go into dev.meta.ai.
So here I am in dev.meta.ai, and I was able to create an API key, and you can already see detailed docs and how you're able to use this in plot code.
And these are all the details that they have provided, codecs.
Then if you want to just call it via Python as well as curl, right?
These are some of the details that have been given.
So what you could do is you can create an API key.
So what I did was, as I mentioned earlier, I'm going to be doing three different demos.
So the first demo here is, think of it like a Facebook Marketplace listing generator.
So here I would be using MuseSpark.
So what I'm going to do here is upload a video here.
So you can see the video very clearly.
This is a video of a bike of a kid where I've just taken like a very short nine seconds.
I just wanted to make it a little bit tough, and I'm going to ask it to generate listing, right?
So there are multiple things which are happening here, right?
So obviously it is calling the MuseSpark 1.1 model.
So the first thing what's happening here is it is going to be identifying the item.
So this is the response back from the first step.
So it has correctly identified that this is a Global Primo Ping three-wheel kit scooter.
And obviously it is used, so it has understood that as well.
So very good job done.
Very quickly, you can see the amount of time it took was very short.
Then the agent is going to now do a pricing research.
So it's going to look into some other prices so that it can give us some comparison pricing, right?
Once you have that idea of the comparison pricing, then it will consolidate all of that and create the listing.
So you can see that.
I'll search current resale listing for your global this.
In the US market, it is around this.
And you can see all the pricing over here.
In UK, it is something of this sort, right?
In Canada, it is this, right?
And because this is coming from Meta and Facebook, so this is Marketplace API.
So you will be able to actually get the ride because they already have the data.
So now it has already created the listing.
So you can clearly see that Global Primo Ping three-wheel kit scooter, hot pink and black.
It also identified this, the best for scooter for toddlers and kids.
This condition is very good.
The retail is for this.
But then here we are charging it at this particular price, right?
So that's what it was able to do.
And you can see that it did a pretty good job.
And you were able to then copy this and just take it and paste it in Marketplace, right?
So the reason I wanted to do this was to show you how you're able to create something like this very quickly using new Spark.
And it really demonstrates an agentic behavioral understanding of a multimodal input in this case video.
And also doing a quick search.
And then not only just limited to the US, but also search across the board and then providing you like a competitive pricing, right?
So I was very impressed with what I saw.
Okay, so that's the first demo.
I hope you enjoyed it.
Now, what I want to do here is I actually want to show you how you are also able to use it directly in CloudCode.
So for that, let's just open Cloud, right?
So in this case, I've already configured Cloud's connector.
So if I show you this, you can see that here I have got the new Spark 1.1 agent here, right?
So the way I was able to do this was I basically followed this specific instruction.
Here I provided the API key, which I already showed you.
And I ran this in PowerShell first, right?
So before actually running Cloud, I basically ran this, right?
So once I have this, now Cloud is being forced to use this particular model, which is new Spark 1.1, right?
So you are welcome to try this and see what kind of results you're getting.
All right, so for the third demo, I decided to actually use one of Meta's cookbooks, which they have given.
And they have really done a great job and provided 10, actually 13 different use cases.
So here in this one, I actually decided to use this perception grounding, right?
So the use case here is you have your fridge filled with some food objects and can Meta's multimodal model be able to get and identify each one of the food objects and provide some sort of a score, right?
So if I go in detail, like this is the original image, you can see all the different food items over here.
And once you basically run this particular program and provide your API key, it should be able to identify each one of these items and provide the score, right?
So what I did was I ran this particular prompt, right?
I'm a pescetarian with high cholesterol, put green dots on recommended food.
And now that it has run this particular output, once I give this particular command, I will be able to see the output, right?
You can see it is now generating the HTML overlay so that it will be able to identify that, right?
So I want to also show you the output of what it produces.
So again, just for the reference, this is how the original picture looks like, right?
So if you look at this is how the picture looks like.
Now, what we will be able to generate is something like this, which is after it has generated the output, right?
So pescetarian plus high cholesterol, fridge guide recommended and non-recommended.
And you can see the scores pretty much well done.
So this is the output.
And I again thought like this was amazing because you can see orange juice, we all think that it is great, but for some reason it is giving not a great score.
So let's see if I eat something which is not recommended.
So cheese pack is not recommended.
Yellow cheese is not recommended.
Whereas this butter is definitely not recommended.
So pretty cool, right?
So you are able to use this out of the box and able to get this label and it's fast, it's cheaper as well.
So this is what I wanted to cover.
The main idea is a make you aware of this is a brand new agentic model out there.
So definitely give it a try.
You can see there like this is a good playground and a dashboard.
So I've been playing with it.
It provides like the usage and stuff like that very well, etc.
All of those things.
And then they have also given like $20 free.
So, so far I've ran it a few times and I only spent less than a dollar.
And there's another way for you to try, which is playground.
So you can upload an image or ask certain questions.
You can also change some of the settings over here.
And then there are some advanced settings as well.
Right.
You can add some JSON schema and stuff like that.
So what I would recommend is give it a shot.
Also try it in combination with Cloud Code.
The app that I built, I asked Antigravity to actually build the app, but leveraging the Spark API.
You could do like all of these types of combination where you use your ID and build an app like this and see for yourself what you feel as the performance.
Right.
And obviously on the benchmarks, they have talked about the benchmarks over here.
They've compared themselves against Gemini 3.1, 4.8.
These are here for your reading.
And I will share this as well.
Of course, it does all different types of use cases from an agentic perspective.
It also does computer use.
It definitely writes code.
All of these things are something which they have actually explained.
Right.
So multimodal is something which we saw live.
Again, hopefully this was helpful.
It added some extra knowledge to your existing knowledge base.
Let me know if you guys have any questions and what you feel after trying this.
Thank you very much for your time.
If you like the video, please hit that like button.
And if you're new here, please hit that subscribe button as well.
Thank you for watching and I will see you in the next one.