Use an MP3, WAV, OGG, or M4A file from 15 seconds to 8 minutes long, up to 50 MB.
Lyric Video Maker for Photos and Your Song
Live preview with audio
An abstract preview will be replaced by generated photos
Upload a track to preview your photo video
Calculated after upload
Choose music to see estimated usage
Sign in before generating a photo video.
Pair your song with photos and captions for a lyric video, then review a local preview before deciding whether to submit a render. A completed task produces a downloadable MP4 with the song audio, selected photos, and captions.
01
Start with an audio file you can use
Choose a song you own from your account’s available audio while signed in. Only completed or otherwise available owned audio is selectable, so a track in your history may not be available for this project.
02
Build a still-photo sequence
Upload one to twelve JPEG, PNG, or WebP images and place them in the order you want viewers to see them. A single image can remain on screen for the whole song; for a changing sequence, use two to twelve images. Think through the progression before rendering: an opening image can establish the mood, while later pictures can mark a change in scene or emphasis without adding motion footage.
Uploaded originals are cleaned after task completion.
03
Add optional AI-generated artwork
AI artwork is an optional still-image source. Describe the subject and atmosphere in a visual prompt, choose one of seven photo styles, and select a standard or refined image-model option. Treat the result as artwork for a photo sequence rather than motion footage. Review each image against the visual direction you have in mind: a prompt guides the request, but it cannot guarantee an exact result or matching character details across images.
This artwork is optional: uploaded images are a separate path and do not require AI image generation. Rendering with uploaded images costs 20 credits. Adding AI images costs 2 credits each at standard quality or 9 credits each at refined quality, in addition to the uploaded-image render cost. Consider how many generated pictures the sequence needs before choosing that route.
04
Supply lyrics and set their timing
Paste lyric lines when you already have the words, import an LRC file when you have timed cues, or adjust caption times yourself. Pasted text without timing is initially distributed evenly, which gives you a starting arrangement rather than a verified match to the singer. Listen and review each cue against the song, especially where a line begins or ends between beats, before you rely on the timing in the finished video.
For eligible songs generated on this site, the tool reuses cached timed lyrics when they are available. Precise new timing costs 5 credits. That option applies to the eligible site-generated song workflow; it does not automatically transcribe arbitrary audio that you upload. If the words or their placement matter, provide the lyrics and check their timing instead of assuming the audio file will be analyzed for you.
05
Choose the frame and caption look
Set the aspect ratio to landscape 16:9 or portrait 9:16 according to where you plan to show the video. A landscape composition gives photos and captions a wide frame; portrait suits a taller display. Review the crop and text placement together, since an image detail near an edge can compete with a caption even when each looks clear on its own.
Choose Clean, Karaoke, Typewriter, or Bold captions and place them at the top, center, or bottom. You can also add a waveform over the retained song audio. Keep captions readable against the image and leave important faces or details unobscured, so text and artwork remain easy to follow together.
06
Review the local preview, then submit if ready
Anonymous visitors can open the workbench, choose local files, and preview the setup before signing in. This local preview is for checking the arrangement; it does not submit a paid server task. Use it to look for an image order that feels right, captions that are legible, and timing that follows the supplied cues. A preview is not a completed downloadable render.
Sign in to submit a server render, then review the estimated credit cost shown before you confirm.
07
Return to the task and download its video
Your render runs as a background task, so you can return to its history to check its status. When complete, it produces a downloadable MP4 with your selected photos and captions plus the song audio.
Completed videos are private and downloaded through the authenticated account route. Sign in to reach the download for your result, then save a copy somewhere you control if you may need it later. The service does not promise that completed files will remain available permanently.
08
Choose the visual format that fits
For lyrics over photos, choose the photo-video tool; for graphics that react to a song, choose the visualizer.
09
If a render is interrupted
What happens if a charged photo-video task fails?
A failed photo-video task that charged credits has a refund path. Check your account history for the outcome.