How do we take sore-l like videos and

post them on any social media and done

automatically? I'm going to show you

step by step how you can create a social

media brand that automatically creates

content leveraging this Sora API.

Welcome back y'all. On today's video,

we're going to be leveraging SOAR API so

we can automatically create video

content. And today's video is sponsored

by Zapier. They hit me up and they said,

Corbin, we have a cool idea for video.

We need to leverage the SOA API. And I'm

like, you know what? You're right. Let's

do it. Zapier is an automation platform

that allows us to connect all these

different softwares you're about to see

in today's video, such as uploading to

YouTube, pushing this data to a Google

sheet, and leveraging the actual Sora

API itself. So, let's go ahead and get

started here. And what I want you to do

is actually create a folder. So, we

create new folder. I'm going to call

this Sora influencer. Save. My objective

is simple. How do we automatically

create Sora videos? And then you can

leverage that and post anywhere on

social media. Therefore, create new zap.

This first automation we're going to do

together is simply going to create the

actual video. It's going to put in a

request and you're going to notice that

it's pending. Therefore, we're going to

create a separate automation to listen

for this request. So, when it's done, we

can do XYZ. So, first things first,

what's going to initiate your creation

of a new video. This in theory could be

maybe you're researching, you find a

cool topic and add information here and

then Zapier automatically grabs this

data and puts it through the pipeline.

In today's video, we'll do a simple

trigger which is going to be a scheduler

schedule. So the logic I want to run in

this pipeline is first we're going to do

a trigger event of how often we want the

pipeline to run. As you can see right

here we can set our frequency. So for

now I will go and say every hour. We

want to put out a new piece of content

every single hour. Hit continue. We can

select the relevant time. So we'll say

every hour at 6:00 7:00 alternatively

6:15 7:15 whether to trigger in the

weekends or not. Continue. The specific

action we're looking for here is going

to be a Sora video generation. By the

time you watch this video it probably

already exists. As of now, we're just

going to do an API request to actually

access this endpoint. So, I'm going do

API request beta. If that option is

still not there by the time you see

this, then don't worry. Just follow the

same steps I'm doing. Select your

account. Continue. Now, this might look

a little confusing. Don't worry. For

one, I'll put the entire documentation

for Sora API in the description down

below. For two, simply hit copy page

here. Paste it into an AI model chat.

Say, "Hey, I want to integrate Sora with

Zapier." It will help you out. It will

show you step by step. For now though,

for us, we're going to do the URL.

Because we're sending data to OpenAI,

we're going to use the method of post.

Think of it like sending mail. You got

mail. That situation. The URL is going

to be this URL right here. The v1/v

videos. This information is outlined

here as well. If you're wondering like

where's he getting all these URLs from?

It's in the docs. Anytime you do a

request like this, we're going to have

to add an additional header here of

content type. This is going to be

application JSON. This is essentially

how we're sending the data. Don't over

complicate this, y'all. If you don't

know what's happening up to this point,

don't worry. Just talk to an AI model or

just paste it in here. All JSON is is

another language. If you speak German,

it's Deutsch Deutsch. If you speak

English, it's English. This is the way

that these little bots and software, it

communicates with each other. Therefore,

JSON in this payload, we need to

identify a couple things. First off, the

model type. This is actually pretty

fundamentally important for you to know.

And what I mean by that is you can see

the relevant pricing associated with

each model here at OpenAI pricing. If I

come down here, you can see the Sora 2

model is only 10 cents per second.

Alternatively, the Sora 2 Pro is 50

cents when you use these dimensions. So,

dive into the documentation a little.

See what your use case is. See which

model is the best fit. The higher level

model will give you higher quality

outputs supposedly. Try it out. Maybe

you like sore too. Next is the prompt.

We all know this. This is essentially

telling the model, "Hey, we want this to

happen in the video. I'm going to say a

cinematic vertical video of a person

walking through a neon lit Tokyo street

at night rain reflecting the lights. I

went to Japan during August. It was

extremely hot. Trust me, there's

summers. You are going to sweat.

Sidetracked. Size 720 x 1280. We want

that Tik Tok vertical format. And then

how long you want it to be. Right now,

we can do 4 seconds, 8 seconds, 12

seconds. For now, I'll do 4 seconds. Hit

continue. I'm going to test this step.

You'll know it's successful when you

type in response here. Or actually,

let's just type in Q. You'll notice it's

successful when you see status is cued.

You run into errors here, you can use

Zapier's AI helper, or alternatively,

use a chatbot. Now, while that video is

cooking, let me show you how you make a

dynamic prompt here if you want to kind

of just push out content on a very

specific niche. We'll do here is

something very special that we've done

for the past couple years here on this

channel. Just add another chat GBT

block, but specifically a block that

identifies a dynamic variable that we'll

use in the next block here when creating

the video. That didn't make sense. Don't

worry, you'll see what I mean. So, I'm

going to choose conversation. I'm going

to continue the model. I'll go GBT5. So,

I went ahead and cooked up a prompt. We

got generate a visually rich cinematic

two-s sentence summary of a random event

happening somewhere in Japan. First term

that is very fundamentally important for

you to see, random. This ensures that

you get randomness in your output. Eg,

one video doesn't match the other video

that came out. So, when you're posting

on social media, you're good to go cuz

it's going to be random. The scene

should feel like it's being captured

through a camera, including framing or

film terms such as wide shot, panning,

close-up, soft light, handheld camera. I

just want to give a bunch of variety

here, y'all. Keep it short, vivid, and

realistic as if describing the opening

of a movie scene. What's big here?

What's your niche? What kind of content

do you want to make for your specific

profile? Is your niche cooking? Or maybe

even niched even further like pasta. Do

you really like pasta? I love pasta. Or

is your niche sports? Or is your niche

fishing? Like you could quite literally

make a whole social media profile where

in the profile it's different people

catching different types of fish. And

it's kind of funny because sometimes the

fish wilds out cuz it's AI generated.

Maybe it's a big bass. Maybe they're

catching like a piece of toast. Who

knows? But the point being though, the

purpose of this is to give us the prompt

for the Sora input. Next, we're going to

put a memory key here. We say Sora

output. This ensures consistent outputs

here. So, we're going to go down and

this should be good and continue. Test

this step. Scroll down. And here's our

prompt. Yshot. Dawn missed over the

Tokyo fish market as forklifts weave

between crates. So on so forth. We'll

actually try this as well so we can see

what it looks like. I'm going to

continue. Continue. And then what we can

do is scroll down to original payload.

And now we're making it dynamic. I

delete this. I add it here. Previous

output, the message content text. Click

that. But what we just created together

is now we have the ability every single

hour to shoot off get a random prompt

but specific to Japan. And then that

prompt is then sent to be created as a

Sora video. So the next question here is

Corbin, how do we get the Sora video?

Where is it? It's in the cloud. But how

do we get to the cloud to get it? Let me

show you. First off, rename this. We're

going to call this Sora Creator. We're

going to back out. I'm in my folder,

Sora Influencer. We're going to create

another zap. So, now that we have a

functional pipeline to create sor

videos, we now need to know when they're

done. So, to do that, we're going to do

a web hook. All I want you to think

about when you hear web hook is this is

our way of listening when the video is

done. So, we're going to do catch hook

here. Continue. Continue. And this is

our URL. Mopy. Because of the fact that

this is using OpenAI's API to create the

Sora video, we need to go back to our

dev portal. This is where you created

your API key. You're going to scroll

down and select web hooks. If you don't

know how I got to this page, you go up

to the projects, hit manage projects.

Once we're here, we're going to create

I'm going to call this zap zap the URL

we just received. And to be honest with

you all, you're learning another lesson

here where essentially you can listen to

any of these different things happening

and incurring within your specific

organization at OpenAI. Now, all we care

about is at the very end here, which is

video completed and video failed. If you

find yourself hitting test trigger and

no data showing up, just go back to the

other automation flow we created

together and just retest step to send

another payload through. What you can

see here though is that it was

completed. Continue selected record.

What we want to do next is add a filter

block. The purpose of this filter block

is going to stop it when it fails. So as

you saw with the web hook we created,

it's video completed and video failed.

Therefore, let's stop the automation

early if the video failed. There's no

reason to keep going, right? So if type

exactly matches video.failed,

the automation would have not continued

as you see right here. Therefore, we

need to make sure it exactly matches

completed because it tells us it

actually worked. Continue. It's golden.

It's green. I love it. So, now that we

have that, we need to get the video

check GBT. JGBT action event. This is

going to be an API request again. Select

your account. Continue. And now due to

the fact that we need to receive data,

grab data. We're going to do a HTTP

method of get and based off OpenAI's

documentation. The way we do that is we

call this URL here v1 videos data ID

content. the data ID because it's

dynamic. We simply add that add and this

will come from the web hook identified

as data ID. Once that's clicked in, now

we'll have the ability to grab the

relevant data of the finished video. I'm

going to go ahead and identify how we

want to accept it. So I'm going say

application octet stream. This is for

video files. Continue and test step. And

yes, the response is a little crazy. So

let me show you a real quick workaround.

We're going to be able to push this

relevant link here and download the

video using Postman. as right now sora

files can be quite large. So it's harder

to pass it through API. So in order to

do this effectively what I want to do is

going to be pass this URL to my Google

sheet. So I'm going do Google sheet here

action event create spreadsheet row.

Select your account. I'll choose my

specific spreadsheet. Aora influencer

worksheet sheet one. Continue. And we'll

go ahead and put in that relevant data

that we care about. Right? So under

download a video come over here. We'll

put request URL. Continue and test this

step. Here we go. Here is the request

URL. I'll zoom out y'all. Now, yes, in

theory, we could remove this step right

here and just grab the relevant data

from here and send it straight to the

spreadsheet row. But I wanted to show

you this step here. So when you want to

integrate it with other platforms like

YouTube or X or whatever it may be

through API, now you know how to grab it

in an automation flow in Zapier. For now

though, what is valuable is we have the

link to the video that exists. So I can

copy this. I'm going to use a platform

called Postman. This allows me to

download directly within the UI. So what

I do is I'll put in that link that we

had earlier. I'll paste it right here.

I'll do a get method. In authorization,

you put in your open API key here. And

when you put your API key here, you hit

send. And here's our video. And before

we play it, let me go ahead and just

download it. So, I'm going to hit

download here. All right. Here's our

video. This is for Tokyo. All right.

This is the forklift prompt. Let's check

it out.

>> True. So, now you know functionally

every single hour a new video will pop

up here for you to download and publish

throughout different platforms. Now, a

really cool workflow you could add on

top of this would be setting up a Google

Drive where every single time you add a

new file in that Google Drive, eg the

video we just created together, it would

automatically upload a YouTube shorts

for example. Now, here's the last two

things I want to leave you with. The

first thing is notice they don't have

any sore watermarks here. So, using the

API, there's no watermark. Second thing,

some of y'all might be like, Corbin,

that's way too short of a video. All you

need to do is come back to this payload.

Instead of four, put eight, put 12. This

is where you identify the duration for

the video. That just about does today's

video. So, make sure to leave a like. It

is completely free. And as you already

know with these style videos, I'll see

you in the next. Zab your plus just

makes it so we can just create content,

content, content, content. So much

content, AI content. I love content.

Zapier plus Sora is the best I've ever

seen.