Descript's captions are a byproduct of its transcript-based editing. You've already corrected the transcript while editing, so captions come out accurate with zero extra work. That's a fundamentally different workflow from tools that generate captions as a separate step. Import your video, let it auto-transcribe, fix any errors by editing text, style your captions, and export as burned-in video or an SRT file.
What you’ll walk away with:
- Accurate captions that were already proofread during your editing pass
- Both burned-in video and standalone SRT files from the same project
- Consistent caption styling saved as a preset across your entire course
- A captioning workflow that takes minutes, not hours, per lesson
Why Descript for captions
Most captioning tools ask you to work on a timeline, dragging subtitle blocks back and forth to match your audio. Descript works differently. It transcribes your video into a text document, and you edit that document just like you'd edit a Google Doc. If a word is wrong, you click it and retype. If a sentence break is in the wrong spot, you adjust it by moving your cursor. It's the fastest captioning workflow I've found for course creators who aren't video editors.
Tools like CapCut also offer auto-captions, and they're great for short social clips. But for longer course lessons — 10, 20, 45 minutes — Descript's text-based editing is significantly faster for catching and correcting errors. You're reading your lesson rather than scrubbing a timeline, and that difference compounds across a full course.
Step-by-step: Adding captions in Descript
Import your video and transcribe
Open Descript and create a new project. Drag your course video into the project, and Descript will automatically transcribe the audio. For a 20-minute lesson, transcription usually finishes in under two minutes. The result is a full text transcript synced to your video timeline — every word is clickable, and clicking a word jumps to that moment in the video.
If you're working with multiple lessons, you can import them all into separate compositions within the same project. This keeps your captioning work organized per course module.
Review and correct the transcript
Read through the transcript from start to finish. Descript highlights low-confidence words in gray, so you can spot likely errors quickly. Common corrections include proper nouns (your course-specific terminology), technical terms, and any words spoken quickly or softly. Your transcript IS your captions, so errors here show up on screen.
A practical tip: if you use the same specialized terms across every lesson (like "pranayama" in a yoga course or "sourdough starter" in a baking course), add them to Descript's vocabulary feature so future transcriptions get them right automatically.
Enable and style your captions
In the Canvas panel (the visual preview of your video), click the "Captions" option to overlay the transcript as on-screen text. Descript offers several caption styles out of the box: standard subtitles at the bottom, word-by-word highlighting (the "karaoke" style you see on TikTok and Reels), and centered text blocks.
For course content, I'd recommend standard bottom-positioned captions with a semi-transparent background. They're readable without pulling attention from your slides or demonstrations. Choose a clean sans-serif font at a size that's legible on both desktop and mobile — 24-32px works well for most video resolutions. Save your style settings as a preset so every lesson looks consistent.
Adjust line breaks and timing
Descript auto-segments your transcript into caption groups, but sometimes the breaks land awkwardly — splitting a phrase across two caption blocks or showing too many words at once. Scan through the video and adjust line breaks so each caption block contains one complete thought. Aim for 1-2 lines per caption with no more than about 42 characters per line. This is the BBC's subtitle guideline, and it's a solid standard for readability.
Export your captions
You have two export paths. For burned-in captions, export the video from Descript with the caption layer visible — the captions become a permanent part of the video file. This is ideal for social media clips and promotional trailers where you can't rely on the platform displaying subtitles.
For SRT files, go to File > Export and select "Subtitles (.srt)." Upload the SRT file alongside your video on your course platform. Most platforms — including Ruzuku, Vimeo, and YouTube — can read SRT files and display them as togglable subtitles. This gives students control over whether captions appear.
Course creator tips
Caption your teaser clips differently than your lessons
For short promotional clips on social media, the bold word-by-word highlighting style grabs attention and works well for the 15-60 second format. For full course lessons, switch to standard subtitle positioning. What works in a scroll-stopping clip feels distracting across a 30-minute lesson. You can save both styles as presets in Descript and switch between them with one click.
Batch your captioning work
Don't caption each video the moment you finish recording. Instead, record and edit all your lessons first, then sit down and caption them in a single session. You'll build momentum with the corrections, your vocabulary additions carry forward, and the style stays consistent. I've watched course creators cut their captioning time in half by batching instead of one-at-a-time.
Limitations
Free plan limits transcription hours
Descript's free plan limits transcription hours, so if you're captioning a large course (20+ hours of video), you'll likely need a paid plan. The transcription accuracy, while good, isn't perfect — you'll always need to review. For courses with heavy technical jargon, multiple languages, or several speakers talking simultaneously, accuracy drops and manual correction time increases.
Overkill for quick social clips
If you just need quick captions for short social clips and don't need the full editing workflow, CapCut's auto-caption feature is faster for that specific use case. And if you need captions in multiple languages, dedicated subtitle services like Rev or Amara offer human translation that auto-tools can't match yet.
Frequently asked questions
How accurate is Descript's auto-transcription for captions?
Descript's transcription accuracy typically lands around 95% for clear audio with a single speaker. Accuracy drops with heavy accents, overlapping speakers, or background noise. You should always review and correct the transcript before exporting captions — plan for about 5-10 minutes of editing per 30 minutes of video. The text-based editing makes corrections fast because you're reading and fixing words, not scrubbing through a timeline.
Should I burn captions into my video or use SRT files?
It depends on where your videos will live. Burned-in captions are part of the video itself — they're always visible and look consistent on any platform. SRT files are separate text files that platforms like YouTube, Vimeo, and many course platforms can display as togglable subtitles. For course videos, I'd recommend exporting both: burned-in for social media clips and SRT for your course platform, where students can turn captions on or off.
Can I customize the look of captions in Descript?
Yes. Descript lets you change caption font, size, color, background, position, and animation style. You can choose word-by-word highlighting (popular for social clips) or standard sentence-style subtitles (better for course lessons). Save your preferred style as a preset so every video in your course looks consistent. Stick with high-contrast combinations — white text on a semi-transparent dark background is the most readable across different video backgrounds.
Related guides
- How to Add Captions Using CapCut — fast auto-captions for short social clips
- How to Edit Course Videos Using Descript — the full video editing workflow in Descript
- How to Use Descript's AI Editing Features — AI-powered filler word removal, eye contact correction, and more
- AI-Powered Captioning Tools for Course Creators — comparing AI captioning options across platforms
From captioned videos to a complete course
Captioned videos are more accessible, more engaging, and more professional — but they still need a home where students can watch them alongside discussions, exercises, and a structured learning path. Ruzuku lets you build your course for free with zero transaction fees — upload your captioned videos, add activities and reflection prompts, and open enrollment whenever you're ready.