Alt text is not a caption track
Alt text describes a still or post. Speech captions transcribe audio. Do not paste one into the other.
Alt text describes a still or post. Speech captions transcribe audio. Do not paste one into the other.
The first carousel frame is a poster. Burn a line into that still. Do not ship a timed track nobody plays.
Story chrome (reply bar, stickers) eats the bottom. Place type higher than you would on Reels.
YouTube chapters are description timestamps. Burned-in lines are pixels. They are different objects.
Spend the first second on the spoken line, not the logo. Burn the claim, then show the pack.
Plain text has no timestamps. An SRT is cues. Burn-in is pixels. Transcribe, then caption — different jobs.
Before you upload: confirm ratio, safe-area, captions, AI label, and SKU spelling on the finished file. The generate is not the check.
Idea Pin safe-area is not TikTok’s caption zone. Recrop the type. A 9:16 Reel master is not a Pinterest-native video pin.
Kling Turbo (12cr, silent), then Gemini 3.1 Flash TTS (4cr), then burn-in. Native audio is wasted when the line must be exact.
A UGC brief is shots, not a vibe: hook line, SKU still, usage-rights window, caption style, and AI-label.
Finish the multi-scene film first. Then transcribe and burn captions. Scene prompts are a bad typesetter.
Captions transcribe 165 spoken-language codes. They do not translate. Dub the locked master, then caption the dubbed track.
CapCut, DaVinci, Opus Clip, VEED, Versely, and Premiere cover junior-editor weeks: cuts, captions, shorts. Pick by footage, not a seat.
Claude, ChatGPT, Jasper, Captions.ai, Versely, and Submagic cover Reels copy: hooks, scripts, on-screen type. Pick by words, not a seat.
A script timer estimates spoken duration from word count. It does not transcribe, time cues, or burn captions into a video.
An AI video editor like CapCut is a timeline. Generate, caption, and publish is a different job. Do not buy one for the other.
A saved kit is injected into generations so hex, type, logo and tone survive a hurried prompt. A PDF style guide does not.
Ship the language set with the first cut: Seedance lip-sync in 8+, 165 caption codes, Octave 2 as an explainer Versely does not sell.
Schedulers will happily post a silent master. Burn captions in the edit, not as a wish in the caption box.
Versely caption families and VEED styles solve the same mute-feed problem with different libraries. Pick by burn control, not by screenshot envy.
Forced narrative, SDH, and CC are three caption jobs. Mixing them is how you ship the wrong file to a platform.
Time captions to the final audio after lipsync. Captioning a pre-lipsync scratch track guarantees drift.
Use the cheap preview path before you pay for a full caption style burn on a long master.
TikTok UI eats the bottom and sides. Caption safe area is not optional if the line is the hook.