Add subtitles and dub your videos

The problem

Video without captions loses a large share of its audience before the first sentence lands, because a lot of viewing happens somewhere sound is not an option. On social platforms the video autoplays silently and the viewer decides in a second or two whether to stay. Without text on screen there is nothing to decide with.

Captions are also an accessibility requirement rather than a preference, and for anything customer-facing that is a reason on its own.

The second wall is language. If your market includes people who work in another language, subtitled or dubbed versions multiply the value of material you already made. Doing it by hand means transcription, translation, timing and often a new recording, which is why almost nobody does it.

What AI can actually do here

Transcription is the foundation and it is genuinely good now, with the caveat that accuracy drops on heavy accents, overlapping speakers, background noise and technical vocabulary. Names and jargon are where it fails most reliably, which matters because those are the words your audience recognises.

Captioning turns that transcript into timed text on screen. For vertical social clips this is close to mandatory, and the animated styles exist because plain subtitles get ignored.

Dubbing is the newer capability: generating a spoken version in another language, optionally in a voice resembling the original. This is impressive and it needs the most scrutiny. The translation can be subtly wrong in ways a non-speaker cannot detect, and a confident wrong statement in a language you do not speak is a real business risk. Have a speaker check anything customer-facing.

Voice cloning carries a separate obligation. Only clone a voice you have the right to clone, and be careful about producing synthetic speech that a listener would reasonably take for a real recording of a real person.

How the tools differ

Descript is the workbench: transcription, editing by transcript, filler word removal and dubbing in one place. Free gives 60 minutes of transcription per month, 100 one-time AI credits and watermarked 720p exports. Hobbyist is $16 per user per month billed annually or $24 monthly, Creator $24 annually, Business $50.

ElevenLabs is the voice engine and is priced on credits rather than seats. The free plan gives 10,000 credits per month with no commercial licence, which is the detail that matters: you cannot use free output in your business. Starter at $6 per month adds 30,000 credits, instant voice cloning and a commercial licence. Creator is $22 for 121,000 credits with professional cloning, Pro $99 for 600,000.

OpusClip generates captions as part of clipping rather than as a separate step, which is the efficient route if short vertical clips are the output. Free has a watermark on animated caption templates; Starter at $15 per month removes it and adds AI voice-overs.

When you don't need a tool

If you publish only to platforms that auto-caption acceptably, check their output first. It is often adequate for casual content and free.

If your videos are internal, captions matter for accessibility but polished styling does not. Do the transcript, skip the production.

And if you are producing the clips as well, handle both in one pass rather than paying twice. See turn one piece of content into a week of posts.