Making a music video usually takes a crew, locations and an editor. Here AI does the shooting and the cutting, and you decide what the video is about.
Upload a song as MP3, WAV or M4A, from 15 seconds to 4 minutes. AI finds the tempo and the downbeats, transcribes the lyrics with word timings (ElevenLabs Scribe) and works out where the verses and the chorus are. GPT-6 then writes a shot list: cuts land on the downbeats, faster in the loud parts and slower in the quiet ones, and the scenes follow what the song is about. Add your own idea, like “two friends driving to the sea in an old car”, and paste the lyrics if you want the scenes to follow the words more closely.
The plan is free. It shows every scene with its time, the part of the song, the lyric and the shot description. Edit any description you like, and the exact price, worked out from the plan, is shown before you start.
Then AI shoots it. Nano Banana 2 draws a frame for each scene, MiniMax H3 Turbo turns it into video, and in the scenes where the singer performs, the lips move to your track. Upload a photo of the singer, plus a second person if the story needs one, and they'll look like the photos. Without photos, AI casts its own characters and keeps them looking the same in every scene. The edit follows the bar grid exactly, and you can add lyric captions in one of five styles.
Pick a look (cinematic, anime, 3D cartoon, retro VHS, neon or dreamy) and a format: 16:9 for YouTube or 9:16 for Reels, Shorts and TikTok. A ~30-second teaser around the chorus costs far less than a full-length video and is perfect for announcing a release. Use your own songs or tracks you have the rights to; photos of celebrities and explicit scenes are not accepted.