How it works

Upload. Transcribe. Caption.

One worker, one queue, no timeline scrubbing. Timestamps are sacred — everything downstream only ever edits text.

Three moves, zero timeline scrubbing.

Drop footage

Any talking video up to your plan's length cap. We probe it, check your quota, and queue one worker — no babysitting.

Words get timed and tuned

Speech becomes timestamped segments. Grammar and fillers get cleaned up, and on paid plans the whole track can be translated — timings untouched.

Captions attached

No re-encode, no waiting. Your original video plus a toggleable captions track, SRT, and VTT — ready in about a minute.

fluxsubs — a typical run

$ fluxsubs run your-video.mp4 --lang es
[captions] 47 segments · timings native
[polish] grammar fixed · filler cut · en → es · timings locked
[captions] track attached · srt + vtt · video untouched
✓ done    fluxsubs_a1b2c3d4.mp4 + subs.srt + subs.vtt

Give every video captions.

Drop footage in the studio. Words land on the beat, captions attach in about a minute.

Caption a video