TL;DR (60-second read): 75% of short-video viewers scroll in silent mode—no captions equals no content. This tutorial walks through a 5-step workflow in CutFast: ① import 9:16 source ② AI auto-transcription ③ style captions (dynamic / karaoke / bilingual) ④ export per-platform ⑤ distribute. All in-browser, no install, no manual typing.

Why Vertical Video Needs Captions in 2026

Context	Mute-play rate	Impact without captions
Commuting	92%	Swipe-away in 3s
Office break	85%	Completion rate halved
Bedtime scroll	70%	Bounce spikes
Auto-play feed	100%	Zero appeal

The old caption workflow—type manually (60 min), align timeline (30 min), adapt per platform—adds up to 2+ hours per short. CutFast compresses it to under 5 minutes, with captions and clip selection coupled: drag your mouse over caption text = the clip gets selected for the final cut.

Step 1: Import Your Vertical Source (5 entry points)

Open cutfa.st and start a new project. CutFast accepts:

Local upload: MP4 / MOV / WebM / MKV, up to 2GB per file
Platform link paste: YouTube Shorts / TikTok / Bilibili / Xiaohongshu / Instagram
Browser screen recording: capture any tab via CutFast’s built-in recorder
Desktop client: safer for privacy-sensitive footage
Cloud drag-in: paste Google Drive or OneDrive links

Gotcha: If your source is 16:9 but you plan to output 9:16, pick 1080×1920 canvas first—CutFast then auto-detects the subject position and suggests the crop. Don’t transcribe in 16:9 then crop, or you’ll lose facial focus.

Step 2: AI Auto-Transcription (No Manual Typing)

After import, tap “Generate Captions.” CutFast’s transcription engine:

Language	Accuracy	Use case
Mandarin	97%+	vlogs, interviews, talking heads
English	96%+	tutorials, global-facing shorts
CN/EN mixed	94%+	tech creators, diaspora
Cantonese	90%+	HK/Guangdong creators
Japanese / Korean	93%+	overseas-facing content

Two actions happen in the caption editor:

Action A: Strip Filler Words

CutFast grays out “um,” “uh,” “you know,” “like” (not deleted, just flagged). Bulk-select to remove—vertical viewers decide in 3 seconds. See AI filler-word removal tutorial.

Action B: Highlight Captions = Select Clips (CutFast’s Core Move)

Unlike CapCut or Descript: drag across caption text, the matching clip is selected for final cut. You don’t cut first then caption—they’re one motion.

Step 3: Caption Style (3 Templates Built-in)

Style A: Dynamic / Pop-up (TikTok / Reels)

Words pop one by one, keyword color shifts
Background: translucent black or colored outline
Font size: 48-72 pt
Position: lower third, 120-180 px from bottom (avoid platform UI)

Use for: entertainment, emotion-heavy clips. Pushes dwell from 1.5s to 5s+.

Style B: Karaoke (Lyric Highlight)

Whole line stays visible; current word changes color
Position: middle-bottom or top (follow subject)
Best for: music overlays, quote cards, teasers

Style C: Bilingual (Essential for Global)

Mandarin on top, English below (or reverse); 20% size delta; color contrast
Needs larger vertical space (roughly 1/4 frame)
CutFast’s auto-translation hits 95%+ on zh↔en

Use for: Chinese creators targeting global, language-learning channels, cross-border commerce.

Size / Position / Outline (cheat sheet)

Platform	Font size	Bottom safe zone	Outline
TikTok	56 pt	180 px	Black 3 px
YouTube Shorts	60 pt	160 px	Black 2-3 px
Instagram Reels	52 pt	200 px	Double outline (black+white)
Xiaohongshu	48 pt	140 px	No outline; translucent bg
WeChat Channels	52 pt	170 px	Black 2 px

Step 4: Multi-Platform Batch Export

CutFast’s “Multi-export” lets you output 5 versions in one click:

Keep 9:16 ratio (fits TikTok, Shorts, Reels, Xiaohongshu, Channels)
Format: H.264 MP4 (max compatibility) or H.265 (smaller, modern phones)
Bitrate: 8-12 Mbps (15-20 Mbps for 4K)
Burn-in vs Soft captions:
- Burn-in: baked into pixels, can’t be toggled off—best for mute-play platforms
- Soft (.srt / .vtt): platform-toggleable—best for YouTube Shorts SEO

Recommended: burn captions into the main export for mute-play retention, plus upload .srt to YouTube so the caption text gets indexed.

Hidden SEO Wins from Captions

Short-video search: TikTok search matches caption text—stuff your keyword into the opening caption
AI-engine citation: ChatGPT and Perplexity crawl caption text when summarizing video—structured captions lift citation odds

Step 5: Distribute (One Cut, 5 Platforms)

From the export panel you can:

Batch-download all sizes (auto-named <slug>-tiktok.mp4 / <slug>-shorts.mp4)
Generate platform upload deeplinks (CutFast doesn’t post on your behalf but hands you fast-upload URLs)
Log publish time & metrics for post-hoc ROI analysis

For batch repurposing a long video into 5-10 shorts, see Long-video-to-shorts repurpose workflow.

Common Pitfalls (Save 30 Minutes Each)

Pitfall 1: AI Mis-Transcribes Proper Nouns

Heavy accents, jargon, brand names drop accuracy to 70-80%. Fix: pre-load a term dictionary in the top-right of the caption panel (brand, guest names, acronyms). AI prioritizes these words.

Pitfall 2: Bilingual Captions Cover the Face

Priority order: ① re-compose subject in the upper 2/3, ② shrink font 15-20%, ③ tighten bilingual line-height, ④ switch to a “side-by-side” variant.

Pitfall 3: TikTok Crops Bottom Captions

TikTok’s bottom 140 px is reserved for follow-button + caption UI. Any caption under 160 px gets eaten. Keep 180+ px safe zone.

Pitfall 4: Karaoke Out-of-Sync with BGM

Lock the voiceover track first, then use CutFast’s beat-align tool to micro-adjust caption in/out points—don’t try to align music to captions.

Pitfall 5: Blurry Captions After Export

Low bitrate. For 1080×1920, ≥ 10 Mbps recommended; enable “subtitle anti-alias” for fonts under 48 pt.

FAQ

Q1: How many languages does CutFast transcribe?

40+ languages for transcription and translation, covering East Asia (zh/en/ja/ko), Europe (de/fr/it/es/pt/ru/pl/nl), and Southeast Asia (th/vi/id). Translation pairs = 40×40 = 1600 combinations.

Q2: Can I batch-process 20 vertical videos?

Yes. Bulk-import, run transcription once, tune styles per clip. Total transcription for 20 shorts: ~10-15 minutes, 2-3x faster than CapCut / VEED.

Q3: Can I mix karaoke and dynamic styles in one video?

Yes. Captions style per-segment, not per-video. Use dynamic for the hook (first 3s), plain for the body, karaoke for quote moments.

Q4: How much free AI transcription do I get?

Free tier: 3 free edits/day, unlimited video length per clip. Heavy users buy Fafa credits ($0.5/min) or lifetime membership ($399 one-time). See CutFast pricing.

Q5: Can I edit captions after export?

Yes. Caption + cut project lives in CutFast cloud (or local file). Reopen, edit, re-export—no redoing from scratch.

Q6: CutFast vs CapCut for vertical captions, which is faster?

CutFast wins on speed thanks to “caption-drag = clip-select”—you don’t edit timeline first then caption. Full comparison: CutFast vs CapCut review.

Q7: Can I export captions as standalone .srt?

Yes. Check “export captions separately” in the export panel for .srt or .vtt download.

Closing: Captions Aren’t Extra Work—They Are the Content

In the old world, captions were a post-production afterthought. In 2026’s vertical video era, captions are the content: they’re the voiceover for muted viewers, the search index for platform discovery, the translation layer for global audiences, the citation material for AI engines. A short-form video without captions delivers 70% less value.

Use CutFast to compress captioning into 5 minutes and reinvest the time where it matters: picking the right topic, editing for rhythm, nailing the thumbnail.

Try it now: open cutfa.st, drop in one vertical video, and experience auto-transcription + drag-to-select in under 5 minutes.

— CutFast Team

How to Add AI Captions to Vertical Video 2026: Complete Guide for TikTok, Shorts & Reels (CutFast Tutorial)