Everything generated fresh
Lyrics, both vocal parts and the instrumental are made for your clip, so no existing recording is involved and nothing gets flagged as another artist's song.
PHOTOS TO AN ORIGINAL PERFORMANCE
Create a short rap performance with your faces, your topic, and new music.
Clear, front-facing adult faces. JPG, PNG or WebP, up to 10 MB.
This song: 10 credits · 10 songs a month for $32.90, or 40 a year for $110
Just want one song?
Short answer: This AI rap video generator takes two photos and a one-line topic and returns a music video of the two of you rapping: original lyrics, two AI voices, a trap beat and lip-sync, rendered by Seedance or MiniMax models. A standard video is 12 seconds, 9:16, 720p, ready in about two to three minutes, with no watermark; settings go up to 1080p and longer songs. One video costs $10.99 ($5.49 for your first); there is no free generation.
Two photos and one idea go in. The words, both voices, the beat and the camera work come back as a single short video, usually in two to four minutes, with nothing to install and nothing to edit.
An AI rap video generator brings several creative steps into one browser workflow. Instead of writing a script, finding an instrumental, recording two voices, and editing a performance, you begin with photographs and an idea. My Rap Duo creates a short two-person video with an original verse, generated vocals, and a new beat. The two people alternate their lines at a shared microphone.
This workflow is useful when the goal is a personal short clip rather than a complete music production. You can make something for a friend, mark a small occasion, or experiment with the visual style of a duet. The tool generates the whole performance. It does not ask you to record a voice sample, learn an editing timeline, or install a mobile app.
The available settings let you control the parts that matter most for a short result. Choose the photographs, scene, topic, duration, resolution, aspect ratio, and video model. Durations run from 5 to 15 seconds, and the default 12 seconds leaves room for a back-and-forth. The Standard, Fast, Pro, and Ultra models cost different numbers of credits. Longer does not automatically mean better. A focused idea is usually easier to understand than several unrelated subjects compressed into one clip.
No. There is no timeline to learn and no vocal take to record. You choose the photos, type a topic of up to 120 characters, then set a scene, duration and aspect ratio. Everything else is generated: the lines, both performances and the shot itself. Most jobs finish in two to four minutes in an ordinary phone or desktop browser, and you can leave the page and pick the result up later under My videos.
Prepare either two individual portraits or one image containing both adults. Face visibility matters more than a complicated background because the generator places the people in a new scene. Use an original, well-lit photograph with a front-facing expression. Confirm that both people agreed before selecting the permission checkbox.
Describe the subject in up to 120 characters. A useful topic names one shared experience or relationship, without private details. Choose the professional studio for a simple performance, or another scene to match the occasion. Set the duration and aspect ratio for the screen where you expect the video to be watched.
Press Make and pick a credit pack: your first song is $5.49, then $10.99 or 10 credits. Follow the job's stages as photos are checked, lyrics are prepared and the video is rendered. Most jobs finish in 2 to 4 minutes, and you can leave the page and return through My videos.
Real outputs of this generator, described in words: what went in and what came out. Each one is the first take, not edited.
In: two AI-generated portraits of fictional adults (no real people), one per rapper; topic: hyping each other up before a big night out; scene: marble lobby.
Out: Both friends trade an original verse at one hanging mic in a marble lobby, lip-synced, with two distinct voices and a trap beat. 12 s, 9:16, 720p, Seedance 2.0, first take, unedited.
Watch the MP4In: two AI-generated portraits of fictional adults (no real people), one per rapper; topic: who said "I love you" first; scene: neon couple.
Out: The couple argue the point in rhyme under violet and coral neon, each with their own voice, on a new beat. 12 s, 9:16, 720p, Seedance 2.0, first take, unedited.
Watch the MP4In: two AI-generated portraits of fictional adults (no real people), one per rapper; topic: who is the worst driver; scene: street cypher.
Out: A back-and-forth roast in a brick courtyard at dusk, one line each, lip-synced to a trap beat. 12 s, 9:16, 720p, Seedance 2.0, first take, unedited.
Watch the MP4Every model the studio offers. Credits are for a 12-second song at 720p, or the model’s nearest quality; the standard model costs 10 credits.
| Model | Quality | Length | 12 s song | Best for |
|---|---|---|---|---|
| Standard · Seedance 2.0 Mini | 480p · 720p | 5–15 s | 10 credits | Quick and clear |
| Cinema · MiniMax H3 | 768p | 5–15 s | 14 credits | Cinematic light, 768p |
| Fast · Seedance 2.0 Fast | 480p · 720p | 5–15 s | 17 credits | Faster renders |
| Pro · Seedance 2.0 | 480p · 720p · 1080p | 5–15 s | 20 credits | Sharper motion |
| Ultra · Seedance 2.5 | 480p · 720p · 1080p | 5–30 s | 37 credits | Best faces and detail |
Prices and limits come from the live studio settings.
Lyrics, both vocal parts and the instrumental are made for your clip, so no existing recording is involved and nothing gets flagged as another artist's song.
Choose 9:16, 16:9 or 1:1, 480p or 720p, and 5 to 15 seconds. The default 12 seconds leaves room for a real back-and-forth between the two of you.
A standard 12-second 720p song costs 10 credits. Faster and higher-end models cost more, so you can trade price against polish on each video you render.
A failed render returns its credits automatically, and a photo rejected during the safety check is never billed as a video. Swap in a clearer photo and retry.
Start with a simple idea that someone can understand without a long caption. A two-person performance works well when the contrast between the speakers is the joke: the early bird and the late sleeper, for example. Keep it friendly and agree together on the final audience.
Try different scenes with the same general theme to see how the visual setting changes the mood. The street scene feels different from a quiet booth, even when both have two performers and one microphone. Each new generation is a separate job with its own credit cost.
Make a video around a milestone that matters to someone you know. A new job, a shared achievement, or a friendship anniversary can be enough. Avoid cramming a full biography into the topic field. A specific, positive idea gives the short verse a clearer focus.
No. The input is one or two photographs plus optional text. The system generates vocals and accompanying music as part of the video, so there is no recording step.
Do not submit lyrics from existing songs. The service is designed to write original material and may replace unsafe or inappropriate topics with a safer, friendly theme.
Choose vertical 9:16, landscape 16:9, or square 1:1. Songs are 720p by default, 480p is available, and the Pro and Ultra models go up to 1080p.
No. Uploading a photograph does not provide a voice sample. Vocals are AI-generated for the performance, and should not be described as an authentic recording of either person.
Songs remain in your private My videos library until you delete them. Open the video detail page to play it or use the download action; the download is HD with no watermark.
Video generation failures refund the job’s credits. If photo approval rejects an image, replace it with a permitted image; the rejected photo is not charged as a video generation.
Bring one photo of each person and a single short idea. Your first song is $5.49, half off your first credit pack, and the 12-second 720p video downloads in HD a few minutes later. Then send it to whoever is in it.
Make our rap video