How to Make Better AI Influencers than 99% of People | Full Course

You've probably seen AI influencers just like Itana Lopez blowing up everywhere.

She looks the same in every single photo, no matter the outfit, the environment, or the lighting. And the

crazy part is she's not managed by some Hollywood level studio or a team of elite AI engineers. It's way simpler than that. You don't need super

than that. You don't need super complicated or expensive AI tools to create an AI influencer just like her.

After building AI characters for months and testing every app and process you can imagine, I found a way that anyone can do this. And that's exactly why I put together this master class so you

can follow a clear step-by-step process to create your own AI influencer from scratch. But before we get into the how,

scratch. But before we get into the how, we need to cover the three biggest mistakes I see people make when they try this. Because these mistakes are usually

this. Because these mistakes are usually the reason they fail. First, they over complicate the process by bouncing between three, four, sometimes even five different tools just to make a single post. Second, when they finally get an

post. Second, when they finally get an image they're happy with, the next one never matches. the character looks

never matches. the character looks completely different and can't realistically be used for the same Instagram account. And third, they end

Instagram account. And third, they end up with overly smooth AI looking results that with today's progress in AI will never pass as a real picture. Even with

all that effort, most people still give up because the monthly costs drain them before they can build any momentum.

After helping dozens of people start creating AI influencers they're proud of, I can tell you that all of these problems are completely solvable, and they can be solved with just one tool.

It's called Open Art and it lets you create consistent, realistic AI influencers without tool hopping, without frustration, and without the ridiculous price tag. So, if you've ever wanted to create an AI influencer that

actually looks like the same person in every post and do it in a way that's simple enough to start today, this video will give you the full step-by-step process to make it happen. Let's get

into it. If you want to follow along, I'll leave a link to Open Art down in the description so you can go through this entire guide with me. When you

first sign into Open Art, you'll land on this homepage right here. For this first step in the process, look to the left side of the screen where you see the image icon. Click that and you'll be

image icon. Click that and you'll be transported into an entirely new workflow. This is where we're going to

workflow. This is where we're going to create the initial image of how we want our AI influencer to look. The very

first thing I want to do is change the model so we get a realistic looking result. If I click here on the model

result. If I click here on the model switch, you'll see a lot of different AI models to choose from. Every model has its own characteristics and style. Some

have a more anime feel. Some are D and D style fantasy art. Others look like 3D cartoons. But for our specific AI

cartoons. But for our specific AI influencer, we want the output to look as real as possible. And for that, Open Art has just released their own AI model built specifically for photo realism.

It's called Open Art Photo Realistic.

You can also use other highquality photorealistic models like Flux Context Pro, Juggernaut Flux, or Flux Context Dev. But for this example, we're going

Dev. But for this example, we're going with Open Art Photo Realistic. Once I

select it, you'll notice the interface changes slightly. That's because this

changes slightly. That's because this model is entirely focused on producing realistic images. Now, we can create our

realistic images. Now, we can create our initial prompt for the character. The

best way to write your prompt is to be as specific as possible. Include as much detail about your character as you can.

Appearance clothing style setting lighting, and camera angle. The more you feed it, the better your results.

Knowing this can be tricky. I've

prepared a ChatGpt prompt template that you can use to create a detailed description of your character. Here's

how it works. Copy the entire template, paste it into ChatGBT, and replace the bracketed rough idea section with your own general vision for the AI influencers look. ChatGBT will then

influencers look. ChatGBT will then expand it into a fully detailed description. My prompt from that process

description. My prompt from that process looks like this. front-facing full body where the whole body of the character is visible in the photo. Shot of a young white woman with shoulder length, softly

curly brunette hair and bright green eyes, natural makeup, slight smile, realistic skin texture, wearing a casual light colored blouse, neutral background, soft even lighting, cinematic detail, ultra realistic

photography style, sharp focus. I'll

copy this straight into the open art prompt window. Once the prompt is in,

prompt window. Once the prompt is in, we're going to improve it even further by enabling the auto-enhanced toggle.

This feature automatically refineses your prompt to match the model you've selected. It's trained on what types of

selected. It's trained on what types of prompts perform best, so it tweaks yours for better results. Now, you could just start with a super basic prompt and rely on autoenhance. Next, let's set the

on autoenhance. Next, let's set the output size. I'll choose vertical 3 to

output size. I'll choose vertical 3 to four for this example, but you can pick any ratio that suits your needs. For the

number of images, I recommend generating more than one for your first try. This

way, you can compare multiple variations and choose the best one for your vision.

Before we generate, I also want to show you a handy trick. If you already have a reference image, whether it's of an existing AI influencer or a real person.

If you remove your current prompt and click image to prompt, you can upload an image and open art will generate a detailed text prompt describing it. This

is perfect for recreating or iterating on an existing look. You can even choose which model it uses to analyze the image for the most accurate description. For

this tutorial, though, I'll stick to my text prompt. I'll paste it back in and

text prompt. I'll paste it back in and click create to generate our initial images. Now that our images are finally

images. Now that our images are finally done generating, let's take a look at them. As you can see, they all came out

them. As you can see, they all came out looking really good, but not all of them match my exact vision. This is exactly why it's so important to generate a higher number of images in your initial run. For example, I really like versions

run. For example, I really like versions two and three here, but the others didn't quite match what I had in mind.

I'm going to select number three as my base image. To download it, just click

base image. To download it, just click the button in the top right corner. Now,

I'm going to take this image and head back into chat GPT. I'll paste the image into the chat window and write a prompt like this. I've created this AI

like this. I've created this AI character that I want to use as a virtual influencer. Can you create nine

virtual influencer. Can you create nine different prompts for this exact same person, but from different angles? Chat

GBT will then give you a list of prompts you can use. Once you have these, copy the first one and head back into open art. In the model menu, switch to Flux

art. In the model menu, switch to Flux Context Max. This is one of the best

Context Max. This is one of the best models for generating multiple angles of the same character. Paste your chosen prompt into the prompt field. Enable

autoenhance and in the image guidance section, upload your original base image, the same one you gave ChatGpt.

With everything set, click create. As

you can see, we now have the exact same character, but from a completely different angle. Repeat this process for

different angle. Repeat this process for all the different angle prompts Chad GPT gave you until you have a total of 10 images. The reason we're doing this is

images. The reason we're doing this is that the AI character creation model we'll be building later needs as many varied images of your character as possible. The more highquality angles it

possible. The more highquality angles it sees, the better it can train and reproduce that character consistently.

I'm going to go ahead and generate every other angle. Now, just copying and

other angle. Now, just copying and pasting the prompts from chat GPT into Open Art. And once I'm done, I'll come

Open Art. And once I'm done, I'll come back with all of my completed images.

And as you can see right here, I now have all of my images fully generated and ready to use for creating our trained AI model. To access AI character creation in Open Art, go to the lefth hand menu and click characters. You'll

see three different options here, and I'll quickly walk you through each one.

The first is start with four plus images. This allows you to create the

images. This allows you to create the most detailed and accurate AI model because it uses multiple reference images. This is the one we'll be using.

images. This is the one we'll be using.

And the second option is start with one image. This lets you create an AI

image. This lets you create an AI influencer from just a single picture.

It's not as accurate as the first method, but it works if you only have one image to start from. The final

option is start with a description. Here

you can simply describe your character's appearance and it will generate a few starting images for you. The reason we didn't go with that last method is because starting with four plus images will always give you the most consistent

results. If you feed it all of the

results. If you feed it all of the images we just generated, you'll get a much more accurate and stable model of your influencer. So now let's click

your influencer. So now let's click start with four plus images. First,

you'll be asked to name your character.

You can call it whatever you want. I'll

go with Brooke. Next, upload all of the assets you just generated into the upload window. Once all the images are

upload window. Once all the images are uploaded, just click create character.

The model will now start generating.

This usually takes about 3 to 5 minutes.

When it's done, you'll get a notification letting you know it's ready, and then we can start using it.

As you can see, our character is now finally done generating. You can find it right here under my characters. This is

what really sets Open Art apart. In just

about 10 minutes, you now have a fully trained AI model that can give you thousands of images, all while keeping your character perfectly consistent with the vision you started with. And the

best part, once it's created, it's yours forever. You can keep using it to

forever. You can keep using it to generate new images anytime you want without retraining it from scratch. So

now, let me walk you through how to use your character. To access it, hover over

your character. To access it, hover over it and click the button in the middle that says create. Once you're in, you'll enter your initial prompting workflow.

Here you'll see three main settings.

They're all on the same interface, but they're laid out in separate sections.

The first is prompt and reference. This

is where you type in your prompt, adjust the relevant settings, and generally provide the creative direction. It works

very similarly to the standard prompt window you're used to. The second is the pose editor, and this one's really fun.

It opens up a 3D editor where you can actively adjust how you want your character to stand. Leg up, hands in the air, head tilted back, whatever pose you want, you can set it right there. The

final option is place character in image, which lets you insert your AI character into any image. Whether it's

something you found online or something generated by AI. For this tutorial, we'll start with the prompt and reference option. Click that and you'll

reference option. Click that and you'll see a more advanced prompting window. It

might look complex at first, but it's actually very straightforward once you understand each section. In the first box, you'll enter your prompt. For this

example, I'm going to place my character in a cozy winter cottage. Just changing

up the environment to show off the model in a new setting. Next, we have prompt adherence. This tells the AI how

adherence. This tells the AI how strictly to follow your prompt. A lower

number gives the AI more creative freedom, while a higher number forces it to stick closely to what you've written.

I usually recommend keeping it between 2 and 3.5 and adjusting based on the results you're getting. Since this is an environmental change, I'll set mine to two. Then we have the character

two. Then we have the character settings, which control how closely the AI sticks to your character's original look. A higher weight keeps everything,

look. A higher weight keeps everything, clothing, accessories, details exactly the same. A lower weight allows the AI

the same. A lower weight allows the AI to adapt elements like clothing to fit the new setting. I'm going to uncheck keep clothes the same so the outfit fits the winter cabin theme and set the weight to 0.5. The image guidance

section is something we'll go over later. Finally, we have the aspect

later. Finally, we have the aspect ratio. For Instagram style content, I

ratio. For Instagram style content, I recommend using a vertical ratio. So,

I'll select 2 to three. For the number of images, I'll choose two and then click create. We'll let those generate

click create. We'll let those generate and then we'll see how they turn out.

And as you can see, the image came out looking really close to our original character. The face is exactly the same,

character. The face is exactly the same, but the environment and all the surrounding details are of course different. This is a really solid

different. This is a really solid result. Definitely something you could

result. Definitely something you could post on Instagram. Now, one of the coolest features OpenR offers is the built-in editor. You can access it by

built-in editor. You can access it by clicking the edit button in the top right corner, which opens a whole new window where you can work on your picture. Besides the standard tools like

picture. Besides the standard tools like cropping, adding text, or removing the background, there are a few really powerful features worth mentioning. The

first one is remove. This has two modes.

Quick erase, which is great for making small adjustments or cleaning up minor details, and magic erase, which is for more advanced removals. For example, if I wanted to remove this bowl in the background of the image, I just select

it. As you can see, the magic eraser

it. As you can see, the magic eraser automatically removes it. It's

completely gone. There's also a creativity level slider. This controls

how the AI replaces the area you remove.

A higher setting will make the AI more creative. Sometimes adding elements that

creative. Sometimes adding elements that don't perfectly match the scene. A lower

setting keeps things more accurate and closer to the original image. The next

feature is the face editor, which lets you change your character's expression.

For example, if my influencer is just sitting calmly, I can make her wink by selecting the wink preset. The AI

processes it, and now she's winking in the image. There are also sliders for

the image. There are also sliders for how intense the change is and even whether the mouth is open or closed. For

example, if I set the mouth open slider to the maximum, it will give me a wide open mouth. As you can see here, it's

open mouth. As you can see here, it's not very natural at full intensity, so it's better to keep it subtle, but it's great for fine-tuning the exact expression you want. The third really useful tool is Impaint. This allows you

to create new objects inside your existing image. Let's say I want to put

existing image. Let's say I want to put a Christmas tree behind my character.

I'd select the area behind her, making sure not to overlap onto her figure, fill in that selection, and then in the prompt box, write a Christmas tree in the background. After clicking create,

the background. After clicking create, the AI generates it. And now we have a Christmas tree in the background. In

this case, it's not exactly the tree I had envisioned, and that really comes down to how detailed your prompt is. But

it shows how you can make big changes to an image without starting over. For now

though, let's head back to our character window and explore the 3D pose editor option. So, going back to our main

option. So, going back to our main screen menu, you can now click on the pose your character option. This opens

up a new interface with the 3D pose editor at the top. When it comes to the prompt here, you want to keep it fairly simple. Don't overexlain exactly how the

simple. Don't overexlain exactly how the character should look because the pose and composition editor will handle most of that. For my example, I'll write

of that. For my example, I'll write something like, "Brooke is standing in a coffee shop in front of a pastry section holding a croissant in her hand. For

prompt adherence, I'm going to leave it at four. For the character weight, I'm

at four. For the character weight, I'm not going to keep the clothes the same.

I'll let the AI choose whatever it thinks fits the scene. Now, we can open the pose editor. When it opens, you'll see a few key controls. At the top right, you can choose the look for your model. Either keep your current one or

model. Either keep your current one or pick from a preset. I'm going to choose the female short preset since my character happens to be on the shorter side. For the pose, you have a big

side. For the pose, you have a big library of presets to choose from. I'll

start with left hand on the hip. As soon

as I click it, the model updates instantly. Then using the small circular

instantly. Then using the small circular handles that appear on different parts of the body, I can raise her right hand so it looks like she's actually holding something. Since in my final photo, I

something. Since in my final photo, I want her to be holding a pastry almost like she's promoting the coffee shop, I'll adjust her arm and hand to match that. You can also resize and reposition

that. You can also resize and reposition the green frame to control the composition. This acts like your camera,

composition. This acts like your camera, so you want to set it exactly how you want the final shot to be framed. For

this one, I'll keep it as a straightforward eyele shot. Once you're

happy with the adjustments, click update pose in the top right. This locks in the pose and sets it as a reference for the generation. From here, you can adjust

generation. From here, you can adjust the pose weight up or down. This

controls how strictly the AI follows your pose. I'll set it to 0.5 for now.

your pose. I'll set it to 0.5 for now.

Select four images and click create.

When the results come in, they look almost exactly how I envisioned them.

But to show you the difference this setting makes, I'll push the pose weight all the way up to one and hit generate.

As you can see, this version sticks much more rigidly to the exact pose, while the lower weight version feels a little looser and more natural. In most cases, you'll want to stay below one so it doesn't look too stiff. This feature is

incredibly useful. It lets you get very

incredibly useful. It lets you get very specific poses without having to write paragraphs and paragraphs describing every tiny detail in your prompt.

Instead, you can just make a few quick adjustments in the 3D editor, and the AI will do the rest. And now let's say you already have an image in mind, but instead of generating a whole new background from text, you just want to

place your character directly into that image. That's where the place your

image. That's where the place your character in image feature comes in.

Once you're back on the main menu, you can now choose place character in image.

This feature lets you take your AI character and insert them directly into any image you upload. For my example, I'll upload a photo taken inside a Louis Vuitton store. Once it loads, you'll see

Vuitton store. Once it loads, you'll see a movable frame representing where your character will be placed. You can drag this frame anywhere in the image, resize it, and position it exactly how you want. In this case, I want a close-up

want. In this case, I want a close-up look. So, I'm going to enlarge the frame

look. So, I'm going to enlarge the frame and bring it closer to the camera, almost at full size, so the character really stands out. Once you're happy with the placement, click save placement. This locks in exactly where

placement. This locks in exactly where your character will appear in the final image. Next, we move on to the prompt.

image. Next, we move on to the prompt.

Just like I mentioned earlier, for the other advanced features, don't over complicate your prompt here. The AI is doing a lot of the heavy lifting and blending your character naturally into the photo. So, if you give it an overly

the photo. So, if you give it an overly complex description, you might actually make the results worse. For my example, I'll keep it simple. I want her standing holding two Louis Vuitton shopping bags, one in each hand, looking serious, and

wearing a stylish, high-end outfit.

That's all I'll write. For prompt

adherence, I'll bring it to 35 so it follows the prompt strictly. For

character settings, I'll toggle keep the clothes the same off since I want her outfit to match the luxury vibe of the store. For the character weight, I'll

store. For the character weight, I'll set it to 1.2 so it keeps our look consistent. And just like before, you're

consistent. And just like before, you're better off generating multiple images so you have options to choose from. I'll

set it to six images and click create.

Once the results come in, you can see how realistic they look. Our character

still matches her original design perfectly. And while there are small

perfectly. And while there are small quirks in some images, the overall look is very believable. This feature is incredibly useful if you want to create realworld scenarios for your influencer.

Whether it's for an ad, a promotional post, or just to make it look like they're traveling or shopping, placing them into real life locations adds a whole new layer of realism to your content. But now, what if you want to

content. But now, what if you want to make a realistic video of your influencer, something that shows them in motion or maybe for a carousel post or a quick Instagram story? For that, Open Art has a feature that lets you create a

video directly from any still image you've generated. Let's take this

you've generated. Let's take this example I just made. I can click into that image and in the top right corner there's a button labeled image to video.

Once you click it, you'll be taken into the video generation interface. On the

left side, you'll see the option to select which model you want to use. This

works with most of the current popular models, but I'm going to choose VO3 because right now it's the best option for achieving high realism. It handles

subtle details and smooth movement really well and is perfect for making our AI influencer feel more alive. Of

course, I'll also show you some results from other models later so you can see the difference. But for now, we'll stick

the difference. But for now, we'll stick with V3. Next is the prompt. For this,

with V3. Next is the prompt. For this,

I'll write something detailed so the AI knows exactly what to create. A

realistic video of a woman from this starting image sitting in a chair, holding a cup, casually, giving a small, friendly wave to the camera with her free hand, then smoothly taking a sip from the cup. Naturally excited emotion,

warm indoor lighting, cinematic realism.

For audio, I'm going to turn it off.

That's just a personal preference here, but in my mind, this is going to be used as a story post, so I'd likely add background music later anyway. For

resolution, I'll select 1080p. For video

mode, I'll go with normal. This gives a slightly longer generation time, but results in a smoother, better quality clip. Now, I'll hit create, let the AI

clip. Now, I'll hit create, let the AI do its work, and we'll check back once the results are ready. Now, let's take a look at the final result. As you can see, the video came out looking very realistic. She's doing exactly what we

realistic. She's doing exactly what we wanted, giving a small wave, making subtle expressions, smiling naturally.

It's the kind of clip you could post straight to an Instagram story without any extra editing. To show you the difference between models, let's try another one called Seance. This model is also focused on realism and is actually

made by the same creators who built Tik Tok's AI models. It has a slightly different style compared to V3. For this

example, I'll use one of the coffee shop images we generated earlier. You don't

have to go through the full process of navigating to it inside Open Art. You

can simply delete the current image and re-upload any image you've already saved. For the prompt, I'll keep it

saved. For the prompt, I'll keep it straightforward. A realistic video of a

straightforward. A realistic video of a woman from the starting image standing in a pastry shop in front of a display case filled with pastries, holding a pastry in her hand, taking a natural

bite, inviting shop lighting, cinematic realism. For settings, I'll choose pro

realism. For settings, I'll choose pro quality mode and keep the duration at 5 seconds. Short, but perfect for a quick,

seconds. Short, but perfect for a quick, engaging clip. Then I'll hit create.

engaging clip. Then I'll hit create.

Once it's finished generating, we can see that while the realism isn't quite on the same level as V3, the movement and details are still impressive. This

is something you could definitely use for Instagram or Tik Tok. Whether you're

showcasing a product, modeling clothing, or creating an ad. Finally, let's try one of the newest models added to Open Art called Wong. One of the great things about Open Art is that it's constantly updated. Whenever a new highquality AI

updated. Whenever a new highquality AI model is released, it's usually added right away, ready for you to use without paying for separate subscriptions. For

this example, I'll take the Louis Vuitton store image we generated earlier and write this prompt, a realistic video of a woman from the starting image standing inside a Louis Vuitton store holding two Louis Vuitton shopping bags

in her hands, gently wobbling them with a happy, satisfied smile. Luxury

boutique lighting, cinematic realism.

The only setting available here is resolution. So, I'll set it to maximum

resolution. So, I'll set it to maximum and click create. When it's done, the result is exactly what I envisioned, a short, realistic video of her proudly showing off the shopping bags, smiling as if she's just made a big purchase.

And that's the point. You can use these clips anywhere in stories, reals, carousel posts, or even ads. The

possibilities are massive. You can take a character you've created from scratch and put them anywhere from the moon to a Monaco beach to a luxury boutique, all while keeping them perfectly consistent for brand deals, promotions, and

storytelling. So now you know exactly

storytelling. So now you know exactly how to create a fully consistent, hyperrealistic AI influencer from scratch, train them in minutes, and generate unlimited photos and videos that look like they came straight from a

professional shoot. And the best part,

professional shoot. And the best part, you can do all of it inside Open Art without tool hopping, overpriced subscriptions, or any complicated setups. It's all in one platform that

setups. It's all in one platform that lets you bring your ideas to life faster than you ever thought possible. So, if

you're ready to start creating the kind of content we just walked through that grabs attention, builds a following, and opens you to brand deals, click my link to Open Art in the description and get started Today.