In this video I show you how to make your first AI video on Higgsfield AI as a complete beginner. No camera, no editing skills, no experience. We start with an empty account, make a character with GPT Image 2.6 (the update to GPT Image 2.5), build a character sheet so she looks the same in every shot, and turn her into a realistic 15-second AI video with sound using Seedance 2.5. Then for a bonus, we drop her into a trending video with Genjutsu.
If you've been wondering how to make AI videos, this is the whole workflow, start to finish. Every image model and AI video generator I use here lives on one plan with one credit balance. Grab my three prompts below, swap in your own character, and make your first AI video today.
⏱ Chapters 0:00 What can a beginner make with AI video today? 1:21 What's the difference between image models and video models? 2:18 Why Higgsfield, and which plan do you need? 4:04 How do you make a character with GPT Image 2.6? 5:12 How does a character sheet keep your character the same? 6:15 How do you make a video with Seedance 2.5? 7:36 How do you change the scene with one prompt tweak? 8:45 How do you put your character into a trend with Genjutsu? 9:40 Is AI video a skill worth learning?
📝 My prompts
Character (GPT Image 2.6 Flare, 3:4): A realistic full-body photo of a 40-year-old woman who runs a food truck, short curly grey hair, round tortoiseshell glasses and a warm smile. She wears a mustard-yellow canvas apron over a white T-shirt, dark jeans and red canvas sneakers, with a small notepad tucked in the apron pocket. She stands relaxed against a plain light-grey background, facing the camera, face clearly visible. Soft, even studio lighting, natural skin texture, realistic proportions, sharp focus.
Character sheet (GPT Image 2.6 Sunburst, 16:9): Using the attached picture as the reference, make a character sheet of this exact woman in three panels side by side: a full-body front view, a full-body side view, and a close-up of her face. Same face, same curly grey hair, same glasses, same mustard apron, white T-shirt, jeans and red sneakers in every panel. Plain light-grey background, soft even lighting, realistic photo. No text, no labels.
Video (Seedance 2.5, 15s, 16:9, 720p): @Image1 is the main character. She slides open the service window of a small teal food truck parked on a quiet city street at sunrise, sets a little chalkboard on the counter, and waves at the first customer walking up. Keep her face, hair, glasses and clothes exactly the same the whole time. One continuous shot, the camera slowly pushing in from across the street. Sound: the window sliding open, a grill sizzling, birds and light morning traffic. No music, no dialogue, no text on screen.
Prompt tweak for the night version: change "at sunrise" to "at night under warm string lights" and "birds and light morning traffic" to "a lively evening crowd".
Keywords: how to make AI videos, AI video for beginners, AI video generator, AI avatar video, Seedance 2.5, GPT Image 2.5, Higgsfield AI, Higgsfield tutorial, realistic AI video, AI character video, text to video AI
That is my first AI video.
I made the person, I made the scene, and I made the sound,
and I didn't touch a camera at all.
10 minutes ago, I had an empty account.
I'm gonna show you every click
from signing up to the finished video,
so you can follow along and make one today.
If you have never made an AI video before,
good, this video is made for you.
So you don't have to have a film degree,
you don't need editing skills,
you don't even have to be good with prompts.
I'll give you all three of my prompts.
They're in the description too.
You can just follow along.
If you run a small business or a side hustle,
this is a fast way to get a video for your page.
You don't want to have to go pay people thousands
and thousands of dollars. You can do it yourself.
We're going to help you get a video created
for your page, your product,
or your ads without filming anything.
We're going to do it in three steps here. First,
we're going to create your character.
Then we're going to make a character sheet so that your
character looks the same in every shot.
It's not a new character when you generate new videos.
Then we're going to turn it into a 15 second video that actually has
sound to go along with it.
And then at the end we'll do a bonus step where
you can put your new character into a trending viral video.
Real quick, there's two kinds of AI tools
that are going to show up in this video.
Image models make pictures, they make images.
Video models make those images move.
So that's the whole difference between them.
On the image side, there's Nano Banana,
which is great for editing a photo you already have,
like changing the outfit or the background.
There's GPT Image, which follows really detailed instructions,
and it's great for product shots and readable text.
There's Soul Cinema, which makes cinematic film-looking pictures.
And then there's Seed Dream, which is the all-rounder for people,
places, and detailed scenes.
On the video side, there's one
which fills in the motion between a first frame and a last frame.
There's Kling, which is good at action and making people move.
And then there's Seed Dance,
which takes your character in your scene
and makes the movement and the sound together for
up to 30 seconds. So today I'm going to be using GPT images
for the pictures because it does exactly
what the prompt says. And I'll use Seed Dance 2.5 for the videos
because it gives the most
realistic results. There are a lot of AI sites out there today.
The reason I'm using Higgs field
is that every model I just named is in one
One place, on one plan, with one credit balance.
So you're not paying for five different subscriptions
that you have to jump around.
And I have a link for Higgs Field,
so I'm gonna open Higgs Field through my link,
which is in the description.
New users who sign up with my link,
they get 50% off their first purchase with Higgs Field.
Sign up here in the top right,
and I'm creating a brand new account right now,
so just so you can see that it's empty,
so put in your email there.
Enter a password. So for this tutorial you need the plus plan
because SeedDream 2.5 is not on the
starter plan. So in plus is 47 euros per month if you
are paying monthly or if you pay annually it's
35 euros per month and that comes with 1200 credits
which is about 600 images or 80 videos
and of course longer or higher quality videos
are going to cost more credits but before you
hit generate it'll always show you how
many credits you're using for that video.
Plus also gets you access to all the models
so you can use whatever you want.
And for 14 days, you get unlimited SeedDream 2.5 at 720p
so you can use it quite a bit.
And for 60 days, you get a whole bunch
of unlimited that they offer for you.
You can pay monthly, you can pay annually,
and of course you can cancel at any time.
Now if you're going to be doing this more professionally,
like making videos for other people as well,
the Ultra Plan might be a better option for you.
Gets you 3,000 credits per month with 1,500 images
or 200 videos on SeedDream.
But to start, you really just need plus here when you're starting out.
And remember, if you sign up through my link,
you get 50% off your first purchase.
All right, so now you have a paid Higgs field account,
and there's a whole bunch of tabs here up at the top.
For this tutorial that we're going through,
we're just going to be using image, video, and then genjutsu.
So you can ignore the other tabs for now.
Step one, we're gonna create the character.
So we're gonna go to image and create image.
Down here is where you select the model
and we're gonna be using GPT image 2.6 flare.
And this is the one that I would use for a character
because it follows instructions really closely.
So if you already have an image like of yourself or another,
you know, person that you want to create an image for,
a character for, you can upload it here.
But in my case, I'm going to describe the person that I want to create.
You can see it's 2.75 credits and I'm going to go ahead and generate.
So my character is a woman who runs a food truck
and I just put everything into the prompt that I want in this image.
who she is what she's wearing the background
the style all that and if writing a prompt like this
seems kind of hard to you you can actually work with ChatGPT
or Claude or Gemini and work with
it to help give you the prompt okay she looks
pretty good i'm gonna generate another one a
couple more options i'm gonna make it three
by four for our character sheet we're gonna make
so that's gonna make it kind of a taller picture
and take out the sides a little bit generate this one
Images generate really, really quickly.
Alright, so this is our food truck lady.
Full body, the face is clear, the outfit exactly like I asked for.
Now the next step is something that beginners usually skip,
and it's why their videos look kinda off,
is we want to make a character sheet for this character,
so that when we're generating videos, it always looks the same.
So a character sheet is basically just a person,
but from different angles,
so that you get all the views of this one person.
So for this prompt, we're going to go to references
and generations, and we're going to use this
lady as the reference,
and I'm going to switch to GPT image 2.6 sunburst, and sunburst is the
version for like precise edits, so it'll keep the person the same.
The aspect ratio for this one,
we're going to do a 16 by 9,
so that's like a normal, you know, computer screen.
And then what we're doing is we're asking
to use the picture as a reference and then make
a character sheet of that exact same woman in three panels,
side by side, full front view,
Full, both sides and a closeup of the face
wearing the same clothes and then we will generate.
All right, so we've got the same woman,
but we got her from the front, we have her from the side,
and then we have a closeup of her face.
So this is our character sheet.
This is all ready to go.
So now it's time to make her moves.
We're gonna go over to the video lab.
The model's already set to C-Dance 2.5, so that's correct.
For references, you can just go back to image generations
and we can use our character sheet
there saying image one is the main character.
She slides open the service window of a small teal food truck.
Her face, hair, glasses, and clothes the same.
And up here where it says image one,
I just type at and then it'll pop up my image here
and I can just click that and that's where that image one comes from.
But the prompt says four things.
Who's in it?
What happens?
How the camera moves and what we hear.
That's really what videos are looking for.
Duration, let's go to 15 seconds.
Does go all the way up to 30
if you're wanting something longer, but we'll do 15.
We're gonna make the aspect ratio 16 by nine.
So that's like a YouTube video.
We'll turn the quality down to 720.
We don't need it, 1080.
You can see it saved some credits.
And again, you can see the cost is right there.
So it's 105 credits to generate this video.
All right, we'll click generate.
One thing I really love about AI
is that you don't have to sit here and wait for it.
You can go work on something else
and then when this is ready, it'll be here for you.
So when it's, so when the video is done, here's what I check.
Is it the same person? Is it the
same face? Does anything look off?
Does anything look AI generated? If it does, I just have to
change the prompt a little bit and regenerate.
But actually this one looked really great just,
you know, first run here.
But the beauty of AI generation is I can change the prompt just a
little bit and get a completely different result.
Let's say instead of at sunrise,
let's say at night with string lights and then instead of birds
and light morning traffic, let's say a lively evening crowd.
So I changed the prompt just a
little bit and watch this. We'll get something
completely different. All right, so let's see
what it came up with this time.
See, so we just changed the prompt
just a little bit and you can
get a completely different result. That's the beauty of AI generation.
I'm going to go ahead and download this one so we can use it.
All right, so we got our character.
We got our character sheet.
We got our first video generated.
Congratulations.
Now the bonus.
We're going to go to Genjutsu,
but you can take a video that is trending on Higgsfield
and put your character into it.
So here are some of the trending videos.
You can pick whichever one you would like.
I'm actually, I'm going to do this soccer one.
I think this would be it.
I'm going to hit recreate.
I'm going to hit the trash can icon.
We'll use our character sheet there.
We're going to switch to 720 quality and plus actually comes with free
genjutsu generation so we're going to use one
of those so this is free this doesn't cost anything
all right let's generate see how
our character looks in that soccer video
You said, love me, take me in your arms, and I have no one left.
And don't let your smell impregnate my arms, and if you abandon me...
So there is our food truck lady playing soccer in that video.
So think about where you were 10 minutes ago.
So you'd never made an AI video.
Now you've watched one come together from an empty account.
You made a character, a character sheet,
a 15-second video with sound, and a bonus trend video.
And this skill is only getting bigger.
Grandview Research expects that AI video generating market
to be more than quadruple by 2033.
So growing about 20% per year.
So this is a skill that you definitely want to learn
and get good at.
And the good news is that anyone can learn this,
even without a background in video,
even without a background in tech.
I just walked you through, it's pretty easy.
So if you want to try it,
sign up through the link in the description.
As a new user, you'll get 50%
off your first purchase on Hicksfield.
Start with plus, and then use your first week
of Unlimited Nano Banana 2 to experiment
as much as you can.
My three prompts will be down in the description as well.
So you can copy them, swap them in,
Experiment a little bit.
Thank you so much for watching.
God bless you all and have a great rest of your day.