Create the face once. Cast it in every post.

Create an AI influencer or your Digital Double. Mention them by their @ in images and videos with sound, make them talk, change the setting and finish the MP4 in the Studio, all without switching tools.

An AI influencer or your Digital Double. Images, video, voice and editing in one place.

See how it works

Who appears

Invent a face. Or bring your own.

Every piece of content starts with who appears in it. In Miraggen, that person is a character you create from scratch or the digital version of someone real: you.

Nina, a fictional influencer created in Miraggen, in a front-facing portrait

AI InfluencerFictionalA character you createDescribe the look, choose the face, approve the body and keep a voice. She exists only in Miraggen.Meet Nina

Official portrait of Junior's Digital Double: a bearded man in a black T-shirt, facing the camera

Digital DoubleReal person, with a declarationThe digital version of a real personA guided capture records face, body, motion and voice. After that, it's you in the scene, no set required.See Junior's Double

Either way, you get an @ to mention in images, videos, speech and edits.

AI Influencer

First, a face.

Describe the look in your own words and pick one of three faces of a fictional person. She gets a full body, a voice made just for her and an @ to mention her in your prompts.

NameNinaWhat do they look like?27-year-old Brazilian woman, golden tan skin, voluminous wavy dark brown hair at shoulder length, brown eyes, full natural brows, subtle freckles on the nose, a wide, natural smile. A selfie taken with the phone's front camera.

Voice

Option 1
Option 2, selected
Option 3

Handle@nina

Interface illustration

Full-body shot of Nina in a white T-shirt and jeans, facing the camera — Generated with AI in Miraggen · Core

Name her and describe her look, age and style
Choose one of three faces, or ask for three more
Approve the full body, made from the face you chose
Listen to three voice options and keep one
Mention her by her @ in your image and video prompts

Memorable characters

Not every persona has to be perfect.

Create characters with exaggerated features and the look of real people: the nose, the ears, the hair, the attitude. Miraggen keeps the look and repeats it in every scene: they dance, talk and step into any video.

Pick up to four marks: nose, ears, hair, eyebrows and more
Or start from an inspiration photo: the character comes out alike, with the marks pushed
The same look in images, videos and when swapping a person

Create a character

Neimadilson: a blond and black bowl cut, a long nose and sticking-out ears — Fictional characters made in Miraggen. The video was made in Core mode, from a description alone.

Ussop: big curly hair, thick eyebrows and wide eyes — Fictional characters made in Miraggen. The video was made in Core mode, from a description alone.

The scenes

One face. Many scenes.

Mention @nina and describe the outfit, the setting and the light. Miraggen uses her reference photos with every request, and each click gives you 1 to 4 images.

Prompt summary

@nina filming a story in her bedroom, window light, white T-shirt

Square, vertical, horizontal, feed and other formats
Reference photos for the product, outfit, setting or style
1 to 4 images per click

Nina films a story in her bedroom, in a white T-shirt, lit by the window — Generated with AI in Miraggen · Core

Nina at an ocean overlook at sunset, with wind in her hair — Generated with AI in Miraggen · Core

@nina filming a story in her bedroom, window light, white T-shirt
@nina at an ocean overlook at sunset, wind in her hair
@nina showing a serum to the camera, bright bathroom, morning light
@nina on a São Paulo street at night, dark jacket, city lights
@nina after a workout in a bright gym, graphite activewear
@nina in a marble hotel lobby, black blazer and silk shirt
@nina in a glass office, beige blazer, daylight
@nina in a dark room with magenta and blue LED light

Digital Double

You in the scene, without building a set.

The Digital Double is the digital version of a real person: you, or someone who authorized you. A guided capture records face, body, motion and voice, and then you just mention your double by its @ to create images and videos with you in the scene.

Guided capture on your phone: point the camera at a code on your computer and follow the steps for face, body, motion and voice
Or upload photos, videos and audio you already have, and complete the rest whenever you like
When the double is of yourself, its voice is made from your own recording, and it says the text you write
Use a reference video and your double repeats the moves: a dance, gestures, a walk, a presentation
You stay in control: a declaration recorded at creation, and the option to delete the double by typing its name

Interface illustration · A real person, shown with their authorization

None of these photos was taken.

All of them were generated in Miraggen with Junior's Digital Double, from text prompts. He is a real person and appears here with his authorization.

Junior's Digital Double in a black leather jacket, leaning on a glass railing at sunset, the city behind him — Generated with AI in Miraggen · A real person, shown with their authorization

Junior's Digital Double with his arms crossed in front of a concrete wall — Generated with AI in Miraggen · A real person, shown with their authorization

The double is created with your declaration that you are the person pictured or have their authorization. The declaration is recorded with the date and is not an identity verification.

Swap the person

Take a video. Swap who is in it.

The place, the camera and the moves stay. Only the person changes: in comes your influencer, your Digital Double or a memorable character, with their own face and body.

Original video

With @neimadilson

The video on the right was made in Miraggen, in Core mode, from the video on the left.

Works with a video you upload or one you made
The people in the background, the light and the framing stay as they were
In Quick generation, the Flow and the Studio

Generated with AI in Miraggen

The video

The image becomes a video. With sound.

Pick an image of your persona as the first frame and describe the motion. Without an image, the video comes from the description alone, without the persona: the image decides who appears.

Prompt summaryThe wind moves her hair and dress. She looks out at the sea, then turns back to the camera with a smile.

First frame: the image you choose
5, 8 or 10 seconds
Sound generated with the video
Vertical, horizontal or square

Generated with AI in Miraggen · Core

The voice

And she speaks.

Your influencer gets a voice designed from a description. Write a short line, or tap Write with AI, and generate a video of up to about 10 seconds, with her mouth following the voice.

Hey, everyone! I'm Lia! Let me tell you a secret: I don't exist. I was made entirely in Miraggen.

Photo, voice and video made with AI in Miraggen, in Core mode. Tap to listen.

Use motion

Found a trend? Your persona repeats the moves.

Upload a reference video and choose who appears: your influencer or your Digital Double. Miraggen generates a new video in which the persona repeats the moves, in the setting from the persona's own photo.

The reference video was also generated with AI, outside Miraggen.

Generated with AI in Miraggen · Core

A dance, gestures, a walk, a presentation
With your influencer or with your Digital Double
The result is a new video: the persona, the outfit and the setting come from the persona's own photo

Setting

Not the right place? Change the setting.

In images, attach the photo as a reference and describe how the place should look: Miraggen creates a new image with the same persona in another setting. In videos, Transform changes the setting with an instruction and aims to keep the person and the movements.

Before: Nina films a story in a plain bedroom — In imagesSame person, same outfit, same pose. Change only the setting: a modern loft with large windows, plants and golden late-afternoon light.Generated with AI in Miraggen · Core

After: the same pose in a loft with large windows and late-afternoon light — In imagesSame person, same outfit, same pose. Change only the setting: a modern loft with large windows, plants and golden late-afternoon light.Generated with AI in Miraggen · Core

In videosChange only the setting: a rooftop terrace at night, with the city lights in the background. Keep the person, the outfit and the movements.Generated with AI in Miraggen · Fast

Also in Transform

Look and outfit
Object
Add to scene
Remove
Lighting

In both cases the result is AI generated and can vary. The new image is a creation based on your photo, not a retouch.

Flow

Show what you have. Connect what you want. Let Miraggen create.

In Flow, everything is a step on a canvas: an image, an audio clip, a description, your persona. You connect one step to another, say what role each connection plays and run it. The result of one step becomes the input of the next.

An animation of the Flow: each step hands its result to the next.

What you bring: image, video, audio, description and person
Steps that create: generate images and video, voice, make it talk, motion, swap a person, edit video and upscale
At the end, the result goes to the Studio

The Studio

Finish where you started.

Bring the scenes, the speech and the music onto a timeline, right in the browser. Create and transform without leaving it: generate images, video and voice, make it talk, swap the person in a clip, borrow the motion of another video, change the scene and upscale. Then add titles and captions and export the MP4, ready to post.

An animation of the Studio: a photo becomes a talking video, with captions, exported as an MP4.

Create with AI: image, video, voice and make it talk
Transform a clip: swap the person, motion, scene and resolution
Video, voice, music, text and caption tracks
MP4 exported in your browser

For people who create and sell

The right creative is rarely the first one.

When content is your living, you're always testing hooks, settings and formats. With your persona ready, every new idea is just a prompt, no set required, and you decide what goes live.

Call sheet

Who	What goes in	Format	View
Affiliates	Your product in the scene with your persona, using a photo of it as a reference	9:16 · image	View: Affiliates
TikTok Shop and social sellers	The product in square and vertical, plus a short clip with sound	1:1 and 9:16 · image and video	View: TikTok Shop and social sellers
Course creators and digital product sellers	A short line from your persona to introduce the offer	9:16 · talking video	View: Course creators and digital product sellers
UGC creators	More scene options for every deliverable, and your own footage alongside what you generate, in the Studio	9:16 · image and Studio	View: UGC creators

Start creating

For brands and agencies

Your brand gets a face. And a voice.

Create your brand's persona once and cast it in every campaign, with the product, the setting and the format each channel needs. Organize by project and deliver clean files.

One persona per brand or client, each with its own @
Projects by campaign, client or theme
Clean download: a brand new file, with no camera, location or software metadata
Interface in Portuguese, English, Spanish, French, Italian and German

Projects

Summer collection@nina
Skincare line@nina
Store behind-the-scenes@nina

Delivery file

Camera: removed
Location: removed
Software: removed
Date: removed

image-7f3a2c91.jpgnew file

Side by side

Two ways to make the same post.

The difference isn't having one more tool. It's not losing the thread between one step and the next.

	The usual way	In Miraggen
The persona	Described all over again in every prompt	Created once and mentioned by her @ in image and video prompts
The steps	Image, video, voice and editing in separate places, with downloads and uploads in between	Everything in one place, and the Library keeps what you make for the next step
The voice	Made in another service and fitted in later, during the edit	A voice created for the persona, with her mouth following the line
The technical choice	A list of models to compare	Three modes: Fast, Core and Max
The final cut	Export and finish in another editor	The MP4 comes out of the Studio, in your browser, as a clean download

Modes and credits

Choose the finish, not the model.

Images and videos come in three modes, and you never pick a model. For images, the cost in credits shows before you generate, and if a generation fails, the credits come back.

Interface illustration

Plans

Start creating today.

Credits every month, room for your personas and everything ready to make images, videos and voice. On the annual plan you pay less per month.

Billing periodMonthlyAnnual

Starter

To start creating with your persona.

Loading the prices

Creator

For people who create often.

Loading the prices

Pro

For people who produce in volume.

Loading the prices

On every plan

For images, the cost in credits shows before you generate.
Failed? Your credits come back.
Clean download: a brand new file, with no camera, location or software metadata
On the annual plan, credits arrive month by month over the 12 months.

Compare the plans in detail

Questions

Straight to the point.

What Miraggen does, in short answers.

Do I need to know how to write prompts or which AI to use?

No. You describe what you want in your own words and choose a mode: Fast, Core or Max. Miraggen decides the technical path for each request, so you don't have to compare models.

How does the persona keep her look from one creation to the next?

With every request, Miraggen uses reference photos of her to keep her look. How many go in depends on the mode: Fast carries less reference, so the likeness is lower. Results can vary between creations, and you choose what to keep.

Does Miraggen verify my identity to create a Digital Double?

No. The double is created with your declaration that you are the person pictured or have their authorization. The declaration is recorded with the date and the terms version, and it is not an identity verification. The guided capture is there for the double's quality and consistency.

Can I create a double of someone else?

Only with their authorization. At creation, you declare whether you are the person pictured or whether you have the authorization and the rights to use their image, voice and identity, and you are responsible for that declaration. For now, the voice is only created when the double is of yourself, and use of a double can be suspended by moderation.

Can I delete my double?

Yes. On the double's page, the workspace owner or an admin deletes it by typing its name. The photos, videos, voice and declaration are deleted from Miraggen, and this cannot be undone. Only the audit record of what was declared remains. The images and videos you already generated stay in the Library until you delete them.

Can I create an influencer from a photo of someone else?

No. The influencer is always a fictional character, created from scratch. For a real person's likeness, yours or that of someone who authorized you, the way to go is the Digital Double.

Can I put my persona in an existing video?

You upload a reference video and your persona repeats the moves in a new video, in the setting from the persona's own photo. Miraggen does not swap the person inside the original video. To change the setting, the outfit or the lighting of one of your videos, use Transform.

What is Flow?

It's a canvas where each image, audio clip, description or persona is a step. You connect the steps, say what role each connection plays and run it: the result of one step becomes the input of the next.

How long are the videos?

Clips of 5, 8 or 10 seconds, with sound, vertical, horizontal or square. A talking video runs up to about 10 seconds. For a longer video, you put scenes together in the Studio.

How does the persona's voice work?

Your influencer gets a voice created from a description: you listen to three options and keep one. Then you write a short line, or tap Write with AI, and generate the talking video with her mouth following the voice.

Will I still need another editor?

Not to assemble and export. The Studio brings scenes, speech, music and text together on a timeline and exports the MP4 in your browser. You place on-screen text with the Text tool, choosing from dozens of fonts.

How much does it cost?

Creations use credits. For images, you see the cost before you generate, and if a generation fails, the credits come back. The Starter, Creator and Pro plans will be billed monthly or annually, and the prices are still being defined.

Does the clean download hide that the content was made with AI?

No. It gives you a new file, with a neutral name and no camera, location or software metadata. When you post, you disclose that the content was made with AI, for example with the platform's own AI label.

What languages can I use it in?

The interface is in Portuguese, English, Spanish, French, Italian and German.

The next scene is yours.

Create your persona. Your next scene starts with an @.

The idea you type stays in this tab and opens on your Home page when you sign in.

Already have an account? Sign in