Trends
Every AI trend, one upload flow
Each trend is a template with its own scene, motion and price. Pick the one that matches the photos you already have, and the rest of the flow is identical.
Hotel Lobby AI
Turn two photos into a trending two-person AI duet — singing into one vintage microphone in a warm orange-lit booth.
2 Photos · 4–15s · up to 2K
Rap Duet
Two photos become a stage rap performance duo video with AI-generated performance audio.
2 Photos · 4–15s · with audio
AI Hug
One or two photos become a warm, natural hug video in the scene you pick — an everyday moment rendered by AI.
1–2 Photos · 4–15s · 14 scenes · up to 2K
AI Rap
One photo becomes a solo AI rap performance video — orange studio, neon street, cyber car drop or stage spotlight.
1 Photo · 4–15s · 3 styles · up to 2K
How to choose a trend
Hotel Lobby AI — the orange-booth duet
The flagship template and the one behind the viral clips: two people stand side by side in a warm orange-lit booth, singing a duet into a single vintage microphone that hangs between them. It is the most forgiving trend for average photos because the framing is a steady waist-up medium shot with even studio lighting. If you only have two everyday phone photos, start here.
Rap Duet — two performers, one stage
Both people perform together on a moody neon-lit stage, trading verses into handheld microphones with generated performance audio. It needs slightly better source photos than Hotel Lobby AI because the stage lighting is moodier, but the payoff is the most energetic clip of the four. Choose it when the two people in your photos would plausibly perform together.
AI Rap — one person, three stages
A solo trend: one photo becomes a full rap performance in the orange studio, on a neon night street, or under hard stage spotlights. Because there is only one subject to keep consistent, it is the most reliable trend for single portraits — and the only one where a pet photo has a chance of working.
AI Hug — the emotional one
One or two photos become a warm embrace in any of 14 scenes: an airport reunion, first snow, a golden-hour street, a birthday surprise. It is the quietest template — no performance, no lyrics — and the one people most often use for memorial and tribute videos.
What every trend shares
The upload flow is identical across all four templates: your photos pass a free readiness check and a safety review, you pick a model (MiniMax H3 at 768P/2K or Seedance 2.5 at 720P/1080P), choose 4, 8 or 15 seconds, and the cost is shown live before you confirm. Finished clips are kept for 14 days with a download link, and credits are returned automatically if a generation fails.
Two ways to pay
One-time credit packs price every template at the same per-second rate, so a pack is the cheaper route once you expect to generate more than a couple of clips. Single-video purchases skip credits entirely — pay once for one clip at a fixed price, no balance required. The live cost counter on each generator shows both options before you commit.
Rules that apply everywhere
Only upload photos you own or have permission to use. No minors, no artist imitation, no copyrighted music. Every upload and every finished video passes an automated safety review, and uploads are deleted automatically after 14 days. The full rules live on the Content Policy page.
Before you generate
Get the input right first
The photos and the scene you pick decide the result more than the trend does. These guides cover the rules that make the difference — and the free checker that catches problems early.
See what each trend produces
Real generated examples for every template — no sign-up needed to browse.
Browse the galleryTips that apply to every trend
Use the newest, least-compressed photo you have. Every template keeps one face consistent through motion — the model can only do that with detail it can actually see. A photo straight from the camera beats the same photo after a messaging app has resized it.
Pick the scene for the person, not just the mood. A stage performance needs a subject who reads well in dramatic light; the orange booth flatters everyone. If one photo is weaker than the other, choose the trend whose lighting is most forgiving.
Generate the 4-second test first. It costs a third of a 15-second render and tells you everything about likeness and scene fit. Re-render only the keeper at the length you actually want to post.
Label AI content honestly. Every platform now has its own AI-content rules — a one-line "made with AI" caption keeps your clip compliant and your audience informed.
Questions people ask before choosing
Can I switch trends after uploading? Yes — your uploaded photos work across every two-person trend, so testing the same pair in Hotel Lobby AI and Rap Duet costs only the second render.
Do all trends generate audio? Hotel Lobby AI, Rap Duet and AI Rap all produce native model audio — the duet voices, the stage performance, the rap delivery. AI Hug is silent by design so you can attach the track that fits the moment.
What if a trend is missing a scene I want? The scene libraries grow with the product — new scenes ship to existing trends before new trends launch, and the pages announce them when they land.