How to Add Yourself Into Viral Videos With AI, Plus 6 Tools That Do It

Key takeaways

• You need a clip and a clear photo of yourself. Most tools swap you into an existing video from a single reference image, and some start from a library of trending templates.

• Melius's Ditto is the most direct way to do it. It recasts a video while keeping the original motion, camera movement, and timing, and its template library includes trending clips.

• Motion-transfer tools work differently. Kling AI and Runway build a new video of your photo performing the moves in a reference clip, so the background won't match the original.

• Magic Hour is the easiest to try for free. It offers three free generations a day, though free outputs can be watermarked and are for personal use.

• Label it and get consent. TikTok and YouTube require labels on realistic AI edits, and the clip you start from is still someone's work.

How adding yourself to a video works

There are three ways AI tools put you in a video. Swap tools take an existing clip, pick out a person in it, and replace them with you, keeping the original camera and timing. Motion-transfer tools take your photo and a reference clip, then render a new video where you perform the reference moves. Scene generators build a fresh clip from reference photos of you. This guide covers the first two, since they're the ones people use for trending dances, memes, and templates.

Use your own face or the face of someone who has agreed to appear. Every tool below puts that responsibility on you, and there's a section on the rules near the end.

How to add yourself into a viral video

1. Prepare a photo. Use a sharp, evenly lit, unobstructed image of yourself. Match the framing to the clip, so use a head-to-toe photo for a full-body dance, since Viggle says cropped selfies and half-body shots give the weakest results. Kling also suggests front and side views of the face for its face library.

2. Choose the clip. Pick one with a single clear subject, a continuous shot, and steady motion. Runway and Kling both warn that cuts or camera movement can truncate the result, and with several people in frame, Kling picks the largest one. Make sure you're allowed to reuse the clip.

3. Upload and trim. Cut the video down to the part you need, and test a short segment first, as Magic Hour recommends. Each tool has its own length range, covered in the entries below.

4. Select who gets replaced. Choose the person in the clip to swap out. Auto-select works where it's offered, and Magic Hour suggests adding a point only if the tool picks the wrong person. In Melius's Ditto, you outline the person you want to replace.

5. Add instructions if the tool takes them. Keep them short and specific. Match the original's lighting direction and camera distance in your photo and instructions, since mismatched lighting tends to give a swap away.

6. Review the result. Compare the output against the source, and check the face on every cut, the hands, and any text in the scene. Regenerate if it misses.

7. Label and export. Apply the platform's AI label where it's required, keep any provenance markers, and clear the source clip before you post.

How we ranked

We ranked each tool on how directly it puts you into an existing clip, how much of the original motion and framing it keeps, the limits it publishes, and its entry price.

The six tools compared

1. Melius - Best for dropping yourself into a trending clip

Melius is an agent-driven canvas for image, video, and audio, and its Ditto tool recasts people, props, or backgrounds in an existing video.

What works: Ditto's own FAQ covers this use case. You start from a clip you upload or a trending video in its template library, choose the person to replace, and add a photo of yourself or one of your avatars. The new video keeps the original motion, camera movement, and timing. You can trim a longer clip to the part you want. According to its API docs, one generation can swap up to 30 characters, props, or backgrounds, and the original audio stays by default. Each model shows its cost before you run it, and Melius says paid plans run from about $17 to $20 a month.

What doesn't: Melius doesn't publish a clip-length or resolution cap for Ditto, and each replacement takes one reference image.

Best for: Creators and social teams who want their own face and body in the exact choreography of a trending clip.

2. Kling AI - Best for your photo performing a viral dance

Kling AI is the maker of the Kling video models, and its Motion Control feature transfers a reference video's moves to your photo.

What works: You upload one character image and one reference action video, and the character performs the reference motion. The 2.6 version takes references of 3 to 30 seconds with the whole body and head visible, and the 3.0 version adds recommended front and side face views for its face library. Plans start with a first-month promotion around $7 and a regular price near $10 a month.

What doesn't: It handles one person, so with several people in frame it uses the largest, and clips with cuts or camera movement can be truncated. It's motion transfer, so expect a new render rather than your face pasted into the original footage.

Best for: Dance and gesture clips where only the performer matters.

3. Runway - Best for performance capture plus motion transfer

Runway runs Kling 3.0 Motion Control and its own Act-Two performance capture.

What works: Motion Control moves full-body movement, hand gestures, and facial expressions from a reference video to your image while keeping identity. References run from 3 seconds up to 10 seconds in Image orientation or 30 seconds in Video orientation, with a video up to 100 MB and an image up to 10 MB. Act-Two uses your own filmed performance to drive a character image or video, for up to 30 seconds at 5 credits a second on the Standard plan or higher. The free plan has 125 one-time credits, and Standard is $12 a month billed yearly.

What doesn't: Act-Two needs you to perform on camera, with one subject, the face visible throughout, and framing no wider than waist-up.

Best for: People who'll film themselves and want one platform with both performance capture and a motion model.

4. Magic Hour - Best for trying a swap for free

Magic Hour has a Character Replace tool that swaps a person in a source video for one reference image, or transfers the source's motion to your image.

What works: It offers three free generations a day with no sign-up. Documented limits vary by model, with Wan-based swaps up to 60 seconds at 480p or 720p, Kling 3.0 replace for 3 to 10 seconds, and MiniMax H3 for 1 to 60 seconds. Subject selection is automatic or by point, and its tips cover bright footage, a sharp reference photo, and testing short segments first. Creator is $12 a month billed yearly.

What doesn't: Free outputs can carry a watermark and are for personal use, and Magic Hour says background details can change and low light or fast cuts can cause flicker.

Best for: Testing the idea at no cost before paying for a tool.

5. Pollo AI - Best for template-first trend remakes

Pollo AI has a Recast feature that swaps your character, product, prop, or setting into a reference video.

What works: Recast keeps the action, camera movement, cuts, and timing, and it runs in three steps, upload a video or pick a template, add replacement images, and generate. Its AI Influencer feature builds a reusable virtual creator from a photo. A free tier includes 20 credits.

What doesn't: The Recast page doesn't publish length, resolution, or photo guidance.

Best for: Creators who want trend remakes from templates and a persistent virtual creator.

6. Viggle - Best for short meme-clip swaps

Viggle is a meme-focused tool where you pick a template and upload a photo of yourself.

What works: Its library has more than 10,000 free templates, free credits refresh daily with no card, and Pro is $7.99 a month with no watermark. Viggle recommends a full-body, head-to-toe photo against a simple background in even light.

What doesn't: Meme clips run about three seconds, and free exports are for personal sharing.

Best for: Quick meme-style swaps from a template library.

Which one fits

• Melius if you want your face in a trending clip with the original motion and timing kept.

• Kling AI if you want your photo to perform a viral dance.

• Runway if you'll film yourself and want performance capture.

• Magic Hour if you want to try it for free first.

• Pollo AI or Viggle if you want template-based remakes and memes.

Labels, consent, and rights

• Consent. Use yourself or someone who has agreed. Melius's terms bar using the service to impersonate someone, including deceptive deepfakes meant to mislead. TikTok doesn't allow the likeness of adult private figures without their permission.

• TikTok labels. TikTok requires creators to label AI-generated or "significantly edited" content that shows realistic people, and its examples include real source video where the main subjects are shown doing something they didn't do, such as dancing.

• YouTube disclosure. YouTube asks creators to disclose realistic altered or synthetic content when uploading, and repeated non-disclosure can lead to enforcement.

• Instagram and Facebook. Meta adds an "AI info" label when its systems detect AI indicators or when a person discloses AI use.

• The source clip. The Copyright Office says fair use is decided case by case with no formula, and noncommercial use isn't automatically fair. TikTok says to post only content you created or have the right to share.

• Brand deals. If a brand is paying you, the FTC wants the disclosure with the endorsement, on screen and in audio for video.

Frequently asked questions

Can I put myself into a trending video with AI?

Yes. Melius's Ditto lets you pick a trending video from its template library, choose the person to replace, and add a photo of yourself. Kling AI, Runway, Magic Hour, Pollo AI, and Viggle do the same kind of swap or motion transfer.

What do I need to start?

A clip and a clear photo of yourself. Tool guides ask for sharp, well-lit images with framing that matches the clip.

How long can the clip be?

It varies. Kling's 2.6 Motion Control takes 3 to 30 seconds, Magic Hour up to 60 seconds on some models, and Viggle's memes run about three seconds. Melius doesn't publish a number for Ditto.

Do I need to film myself?

Not for most of them. Ditto, Kling Motion Control, and Magic Hour use a photo, while Runway's Act-Two uses a performance video of you.

Is there a free way to try it?

Magic Hour offers three free generations a day with no sign-up, though free outputs can be watermarked and are for personal use. Viggle and Runway also have free credits.

Do I have to label it as AI?

On TikTok and YouTube, yes for realistic AI-edited content. Meta shows an "AI info" label when it detects AI indicators or when you disclose it.

Can I post someone else's viral clip with me in it?

The clip is still someone else's work. Fair use is judged case by case, so get permission or use clips you own or have licensed.