Become any character. In minutes.
See yourself as a character in under a minute — any selfie, first portrait free, no account. Then bring it to life in about 15 seconds, or record a clip and have your character copy your exact moves. Your progress is saved in this browser, so you can come back anytime.
One photo in, a character video out — made with these five steps.
Step 1 — Upload a photo of yourself
A selfie or a photo someone takes of you, standing where you plan to film. In the next step you become the character, but everything else stays as it is here: the room, the lighting, and how you are framed. Same spot, same outfit, same light as the clip you will record later.
JPG, PNG, or WebP · at least 300×300px · up to 10MB
▸Photo tips — 20 seconds, better results
▸Max resolution, clean lens, no blur
Wipe the lens, hold steady, shoot at the highest resolution your camera offers. Grain in the photo becomes grain in the video.
▸Soft, even lighting
Window light or a ring light. Avoid harsh shadows across the face — the AI struggles to reconcile them later.
▸Plain, uncluttered background
A simple wall beats a busy room — less noise means a more faithful scene.
▸Same outfit, spot, and light you will film in
The photo must match the reference video you record in step 4. Easiest trick: record the video first and screenshot a frame as your photo.
▸Pro tips for a clean result
⚠️ Mistakes that scream “AI-generated”
📐 Mismatched camera angles
Different angles between portrait and video force the model to invent geometry, which shows up as warping and glitches.
💡 Different lighting or location
The model cannot reconcile a bright video with a moody portrait — you get flicker and colour shifts it will never average out.
🖼️ Low-quality character images
Blurry or pixelated references produce muddy, distorted faces every single time. Quality in, quality out.
🌀 Overcomplicated movement
Slow and deliberate always wins. Fast gestures overwhelm motion transfer and are the fastest way to look obviously AI.
✅ Polish that sells the result
🔄 Regenerate, don't patch
If something looks off, adjust the inputs slightly and rerun. It is faster than fighting a bad render and always ends better.
📸 Fix the source, not the settings
Consistently disappointed? The problem is almost always the input images. Re-shoot your photo or find a sharper character reference.
📝 Tweak prompts, don't rewrite
The merge templates are tested. Make small adjustments for your scene rather than starting from scratch.
🎬 Expect one or two attempts
First renders often need a small input tweak. Most people land a clean result by the second try.
Frequently asked questions
▸Is Vidshift free to try?
Yes. You can start without an account — your first portrait merge is free — and creating a free account adds 35 credits, enough for another merge plus up to 8 seconds of animation at 720p (with a watermark). No card required.
▸Do I need permission to swap someone into a video?
If they are a real, identifiable person — yes, always. Swapping a real person into video without their consent is a deepfake and is banned under our acceptable-use policy. Fictional characters, AI-generated characters, and your own persona are what Vidshift is for.
▸How long does a character swap take?
The portrait merge usually finishes in under a minute, and the instant clip that brings it to life takes about 15 seconds — so your first video is roughly two minutes from selfie to playback. Having your character copy your own recorded moves is the heavy step: typically 4 to 7 minutes, plus the time to film the clip.
▸What does the reference video need to be?
MP4 or MOV, 3 to 30 seconds, up to 100MB — a normal phone recording works as-is. Film yourself performing the motion you want, in the same spot and framing as the photo you uploaded in step 1.
▸Does it work on a phone?
Yes. Everything runs in the browser with no app to install, and progress is saved per browser — so you can shoot photos on your phone and finish there, or start on desktop and upload the video from your phone later.
▸What resolution can I generate?
Free accounts generate at 720p with a watermark. The Creator plan ($19/month) unlocks 1080p and removes the watermark; Pro ($49/month) goes up to 4K. Studio generations with Veo 3.1 include audio.