Make a reel with video, music and timed captions

Arrange your story

Add up to eight video clips. Reorder and trim them, adjust speed and frame each clip for portrait video.

The selected video ranges, after speed changes, must total at most 120 seconds. Portrait framing can fill the frame or show the complete clip with white margins. Adjust the focal position when cropping.

Mix original sound and music

Add optional music. Choose the source range, reel start time, looping, fades and volumes. Enable ducking to lower music during original sound.

Music can begin later in the reel, repeat to fill the remaining duration, or play once. Automatic ducking follows the original audio’s activity, rather than recognizing speech. A linked limiter protects the final mix from clipping.

Caption and review

Add short timed captions, inspect platform safe-zone guides, then prepare a 10-second encoded sample or export the complete reel and cover image.

Caption times refer to the joined reel timeline. Reordering or retiming clips does not automatically move captions. The 10-second sample uses the same rendering and mixing path as the full export. Check the actual encoded result before sharing. A cover PNG and optional SRT text export help prepare your post.

Limits and privacy

The finished reel is limited to two minutes (120 seconds), with up to eight video clips, 12 non-overlapping captions, and one optional music track. Source media can be up to 100 MB combined; music is capped at 100 MB. Individual sources can be up to 10 minutes, but only selected video ranges count toward the reel. Select up to 120 seconds of music. MP4 export uses H.264/AAC at 720 × 1280 or 1080 × 1920, with peak protection. Clips join with clean cuts. Captions are burned into the video; text wrapping supports up to four lines per caption. They are manually entered, not automatically transcribed. Original preview shows the selected source clip and guides; prepare an encoded sample to hear the actual mix. The sample contains the first 10 seconds, or the full reel if shorter. Ducking responds to original-audio volume, including background sounds; it does not recognize speech. Repeated music has simple loop seams. Platform guides are approximate and are not exported; interface positions change. No platform publishing, music library or licensed tracks are included. All files stay on-device. A pinned FFmpeg WebAssembly engine downloads from jsDelivr on first use (about 31 MB); internet access and a modern WebAssembly browser are required. Rendering can take time on phones. Cancel terminates the worker. Export re-encodes and cannot recover missing quality; download before leaving the page.