Transcription
Hey Tiffany, what are you doing around here? You know exactly what I'm here for. I want to make AI influencers for free. You came to the right place. Meet Tiffany. She has never held a camera, but she already has an audience, a personal brand, and famous brands lining up to work with her. And the crazy part? She is not real. Or is she?
There's a shift in the social media space, and anyone with a laptop can build an AI influencer just like her. How to achieve perfect consistency with completely free AI tools and how to actually make money from it. What are we waiting for? Let's get started now.
The way we wanna create our animated AI influencer is by first making sure we have a solid workflow for generating consistent images. For that, we need a clean character board of our AI character. Once you have that, you simply use it as a reference image every time you generate new shots of your AI influencer, and you will achieve stunning results that are both realistic and incredibly consistent. Let me show you how to create one for free.
I'll be using Nano Banana 2 inside Google Flow, which is completely free. And by the way, I'll be covering the best free image and video generation models available right now, all in this video, so make sure you stick around for that.
The first thing I'll do is head over to Pinterest for visual inspiration. I'll scroll through until I find an image that catches my eye and matches the vibe I'm going for. I really liked how this one looked, so I'll just save it to my PC. From there, head over to Chat GPT or Claude AI. Now, I've already prepared a Google DOC with every single prompt I'll be using throughout this video. It will be linked in my Telegram channel, and the link is in the description. Inside that DOC, you'll find the character extraction prompt. Copy it, paste it into the prompt box, and hit generate. Now, this chat is fully optimized to output Nano Banana 2 prompts. From there, we just need to upload the image that we got from Pinterest, and Claude will analyze it and craft the perfect prompt that will produce a similar but completely new and unique character.
Now, head over to Google Flow, which is the official source of both the Nano Banana image models and the Vo 3.1 video model. You can generate unlimited images completely for free, and you get enough credits for up to two videos per day. Right here at the bottom, you can select the model you wanna use. I'll set it to image mode and select Nano Banana 2. I'll set the resolution to 9:16 and the number of images to 4. Now, I'll simply copy the prompt we generated inside Claude and paste it directly into the prompt box. Let's generate a few times and see what we get. So, we've got a few solid results here, exactly what we wanted.
Now, we also need to generate multiple angles of this character so we can see her in different lighting conditions, different facial expressions, and different emotions. Let's go ahead and do that right now. First, we'll set this image as the reference image, and here we're going to say, "Generate a side profile of this woman. Keep it nice and simple." Then, "Generate an image of this woman from the back, focusing on her head." And then we can also go ahead and generate a close-up of her eyes so we can capture the fine details of her face up close. And look at this, we're already getting multiple angles, and we're even picking up even the smallest details.
So now we've got two side profile shots, a back-of-head angle, and close-up eye shots, which gives us incredible flexibility for generating detailed facial shots later on. But we also need a full-body perspective. We need to understand what her complete look is, the full outfit, the proportions, the height. So, let's prompt: "Generate a full-body shot of this woman wearing a black sleeveless crop top and matching high-waist black fitted leggings, dark sneakers. Make sure it is on the same white background." This gives us outfit consistency across every scene, and it also gives us important context around her overall frame and height.
Now, to make sure the outfit stays consistent across all the other angles we already generated, I'm going to upload both images together and do a clothing replacement. I'll prompt: "Make her wear the black outfit from the full-body image in the close-up shot." This keeps everything visually unified across the entire sheet. So now we've generated everything we need. From here, I'll simply download the images. When you hit the download button on a free plan, you only get the 1K resolution option. If you're on Pro, you can upscale up to 2K, and if you're on Premium, you can go all the way up to 4K. But don't worry, later in this video, I'll show you a completely free method to upscale your images without spending a single dollar.
Now, we're going to take all the images we downloaded and combine them into one single collage. This will be our master reference image, or a character board, in other words. So, whenever we generate a new scene and upload that one image into Nano Banana Pro as a reference, every single shot we produce from that point on will automatically pull from all those different angles and details of our character.
I'm going to jump into Picsart and show you how to create a clean character board just like the one I put together for my AI influencer. You can use Canva or Photoshop as well. On the left side, click on "Start New Design" and select "Create from Blank Canvas." From there, adjust the custom dimensions, type 1920 by 1080, and click "Add." Now, once you have the project created, simply select the "Upload" tab and upload the images. Now, start pulling all of those images onto the canvas. Sizing doesn't need to be perfect, but you do want your main portrait to be large enough so that when you upload it into the prompt bar, the model still has enough visual context to work with. What I like to do is to position the main character image first, full size, front and center, and then arrange everything else around it: side profiles, close-ups, detail shots, build it out from there. That way, when we upload this character sheet into Nano Banana Pro, it has full context: front profile, side profile, facial details, outfit, everything. I'm going to scale up the eye close-up as well so the model picks up on those fine facial details. So now we've got this beautifully cluttered but fully organized character sheet. Export it, and we're ready to go.
From here, just as I said, use this image that we created as a reference image in all your image generations, and you will achieve perfect character consistency, no matter which angle you generate. We can generate images purely from a text prompt using our character reference image, just like I did here. We can also seamlessly drop our consistent character into any background or location we want, just like in this example. We can even swap the whole character out completely using basic prompts and the reference image we built.
Now, let's cover all the best free models available on the market right now, because let's be real, that's the real reason you clicked on this video. For image generation, my primary choices are Nano Banana Pro and Nano Banana 2. I use those two image models for 90% of my generations, and you can use them completely free inside Google Flow. They are perfectly optimized for AI influencers and deliver the best realism and overall quality. Chat GPT is a great option as well. It shines when it comes to image editing. It changes only what you ask for while keeping the lighting, composition, and character likeness perfectly consistent. That alone makes it incredibly useful for refining your shots without breaking your character.
Pro tip: Images are 80% of the job when it comes to creating a great AI video. So always make sure you're working with maximally upscaled, highest resolution images possible. This is what separates clean, cinematic shots from blurry, low-quality outputs. Always upscale your initial images to at least 2K resolution. That is sufficient for almost every major video generation model out there. If you do a quick Google search for "free image upscaler," scroll down slightly, and you'll find the website I use right here. Simply upload the image you want to upscale and select between 200% and 400% enlargement. For our shots, we'll go with 400%, so I'll leave that selected and hit upload to start. Within just a couple of seconds, we get our image back at double the resolution we started with. It does make faces look a bit plastic, so in some cases, it works great, but in other cases, it can ruin the integrity of the image and make the face look too polished.
Let's talk about videos now. For talking videos with the best lip sync, use Vo 3.1 on Google Flow. You get 50 daily credits, which is equivalent to two videos per day for free, and you get an additional 100 credits when you first sign up. It's really simple to use. Just select "video" from the menu at the bottom, set the resolution to vertical, and set it to one video per generation. At the bottom, you can choose your generation model. V 0 3.1 fast is the way to go here. It's cost-effective and performs incredibly well. Now, I'll simply upload the reference image I created earlier. In the prompt box, I'll type out a clean, simple prompt: "She talks naturally and realistically with an American accent. She is in a podcast environment and is not looking directly into the camera, as if she is speaking to another person. She smiles at the beginning and she says, 'It is unreal that even with free AI tools, this looks so real.'" Now, hit generate and let's see what we get.
"It is unreal. Yeah, that even with free AI tools, this looks so real." And honestly, it looks pretty good. Nothing weird happening, no distortion. She looks realistic, natural, and completely believable. It does add a watermark, but that's an easy fix. You can remove it using a site like vmake or clean it up manually in After Effects or CapCut. Simply zoom in slightly and reposition the frame so it's completely out of shot, or use the AI remove feature inside CapCut if you have a Pro version.
Now, here's a gem I found. It's called Arena AI. It is completely free as well, and it works with both image and video generation. To access this, just head over to Arena AI, and at the bottom, you will see different options to choose from. Simply select "image" or "video," upload your reference image, type in your prompt, and start blasting through your generations. It runs on the Grok video generation model, and there is also LTX2, which does not output nearly as good results as Grok. The lip sync isn't perfect. Let's talk about videos now. For talking videos with the best lip sync, use Vo 3.1 on Google Flow. But after enough generations, you will land on a clip with decent sync. Let's talk about videos now. For talking videos with the best lip sync, use Vo 3.1 on Google Flow. And where Arena AI really shines is scenes without dialogue, zoom-ins, transitions, movement shots. This model handles all of that better than anything else out there.
Now, if you want to swap your character into a video and copy the exact movement from a reference clip, you need one character swap. Simply head over to one.video and click "Try Now." This will take you straight to the main generation tab. At the bottom, you can select different generation modes to do a character swap or animate a photo. Select "Avatar," and from there, choose either "Character Swap" or "Photo Animate." Both work on the same principle. Just upload the reference image of your character and the reference video you want to mimic. From there, just set the generation mode to Pro and let it run. As you can see, when I try to generate, it shows zero credits used, and you get 10 free daily credits automatically. So, in practice, it's completely unlimited. The one downside is that the waiting queues can get quite long, and the wait times can add up. But considering it's entirely free and unlimited, that's a trade-off I'm more than happy to take.
One more free and unlimited platform I found is called Q. When it has both video and image generation, and it works really solid. It is a bit slow for video generation, but who cares? It is free, and you can generate as much as you want.
Now, let's talk about the monetization side. For a long time, influencers were just influencers. They would post good-looking photos, show cool places they visited, and share pieces of their lifestyle. And for a while, that was enough to build a huge audience. But social media has changed. Today, people don't just follow someone because they look good or live an interesting life. Most people follow accounts based on their interests, things they want to learn about, or topics they actually care about. And that shift applies to AI influencers too.
There are actually two types of influencers nowadays. You have basic AI influencers like Aitana Lopez, Lil Mckella, etcetera, who mostly post lifestyle-style content. And then you have niche AI influencers who focus on a specific topic and build an audience around that interest. And in today's world, those niche AI influencers are often the ones growing the fastest. This is Rabbi Goldman, and as you can see, he's sitting at 1.4 million followers with videos consistently pulling six figures, sometimes even seven-figure views per post. He will have much more followers once this video comes up, so make sure to check it out. It's insane how fast he is growing. But here's what most people scrolling past don't realize: he is not real. Rabbi Goldman is a fully AI-generated character, and nobody seems to care. Or maybe they do, and they follow him anyway. The character leans into one of the oldest and most trusted archetypes in human psychology: the wise elder, a Jewish rabbi figure radiating authority and timeless wisdom. That positioning alone builds an almost instant layer of trust with the audience, and that trust, he monetizes it with a single nine-dollar PDF in his bio. A guide on how to make money. That's it. No complex funnel, no subscription, no high-ticket course. Just a nine-dollar entry point and over a million people who believe in what he represents.
But what if you don't have a product to sell, and what if you don't have a huge following yet, but you still want to monetize your AI influencer? Sure, as your Instagram grows, you'll eventually land brand deals and affiliate partnerships, but I'm not talking about that. I'm talking about real, recurring income, the kind where people subscribe and pay you every month to follow your content. There are a few options that most of us already know about, but I found something different. A platform that's completely private and fully built for AI character monetization. That platform is called Room 11. I'm going to create an account right now. It takes less than a minute. I'll drop the link in the description if you wanna follow along.
First, log in using your Gmail account. Once you're in, the first thing you need to do is select your birthday, then type in your username, and add a profile image, but I will skip the image for now. The last step is selecting the "Join as Host" option, and that's it. You now have your own private room set up. From there, just accept the terms of service, connect your social media profiles, and do an identity verification to finish the process.
Now, here's the reason I actually care about this platform: the payout rate is 90%. The platforms that pretty much everyone starts on for fan subscriptions, they take 20% of your total earnings, leaving you with only 80%. Room 11 takes just 10%, and on top of that, you get paid daily, not in two weeks, not whenever the platform decides to process it. The very next day, the money is in your account. If you're doing any kind of volume, like running two, three influencers, that difference adds up insanely fast.
Now, here's where it gets interesting. Room 11 comes with a free built-in CRM. It's called Concierge 11. You can manage multiple AI creators from one place, track stats, manage staff, everything. Standalone CRM tools that do this stuff charge anywhere from $200 to $600 a month per page. A CRM basically works like your command center. It keeps track of messages, subscribers, performance stats, and your entire workflow in one dashboard. This is built-in, completely free from day one. If you're planning to run more than one AI influencer, which honestly, you should be, this alone saves you hundreds of dollars a month.
This next one, I haven't seen on any other platform. They have a fully managed chatting service built-in, meaning a real team handles your fan conversations for you 24/7. No hiring chatters, no managing people, no headaches. They take a cut, you keep 60% overall, and you can cancel monthly with zero lock-in. Their renewal rate on this service is over 90%, so it clearly works. For an AI influencer setup specifically, this is huge because you're not sitting there responding to DMs manually. The whole thing runs while you sleep. And this part is directly for people watching this video: Room 11 has full support for AI models. They built this with AI creators in mind. They have AI video generation tools coming soon, integrated directly into the platform at the lowest price available. So everything we built today, the AI influencer, the content, you can plug it straight into this.
But there's one more thing I want to mention before I wrap up: traffic protection. Every fan that comes through your link sees only your content. No discovery pages, no competitor suggestions, nothing. Your audience stays yours, completely. So that's the full picture. We built the AI influencer, we set up the content workflow, and now you know exactly where to take it to actually monetize it properly. Now, go hit Instagram and TikTok with your content and have your Room 11 profile set up from day one, so every viewer that lands on your page has somewhere to convert. Room 11 is completely free to start. Link is in the description. If you're already on another platform, they handle migration, so it's not a big move. If you found this video useful, drop a like. It genuinely helps. And remember one important thing: the future is in your hands, and you are the one generating it.