Which second caption language for a Reel? Read the Audience tab
Edits shows where followers are and offers bilingual captions in 15 languages. Choose the second language from audience data, and burn cues for the rest.

Pick the second language from the largest follower location that is not already covered by your spoken language, then check that Edits offers it. Audience insights and bilingual captions both live in the Edits app, so the data and the tool sit together.
NapoleonCat's roundup of 2026 Instagram updates lists Edits audience insights, including follower demographics, alongside the AI assistant and desktop version (NapoleonCat, read 2026-10-07). Storyboard18 lists the 15 caption languages: English, Spanish, Hindi, French, German, Japanese, Korean, Portuguese, Bengali, Gujarati, Kannada, Italian, Thai, Indonesian and Russian (Storyboard18, read 2026-10-07).
A four-step choice
- Open Audience in Edits and note the top follower locations and what the audience is interested in.
- List the languages those locations use that you do not already speak on camera.
- Check each against the 15 supported languages. If one is in, use Edits' second track.
- If a language is not in the 15, burn translated captions yourself, as the table below shows.
What changes by case
The right tool depends on whether the language is supported in Edits and on who translates.
| Case | Route | Sume field if burning |
|---|---|---|
| Language is in Edits' 15 | Edits second caption track | Not needed |
| Language not in the 15 | Translate yourself, burn cues | cues with text, start, end |
| Spoken language differs from default | Give speech-to-text a hint | language (for example en or ko) |
| Korean on screen | Hangul style | style black-outline or another Hangul style |
Making a translated cue sheet
Get timings from the speech first. Video inspect with transcribe: true and segmentation.mode: "sentence" returns sentence segments that fit caption lines; speech-to-text is $0.01 per audio minute plus compute, and language_code is a hint only. Have a native speaker translate each segment, then send the lines as cues to video captions ($0.20 per job for up to 60 seconds).
The language field on a caption job tells speech-to-text what to expect. It never selects the style or the font, so a Korean cue still needs a Hangul style.
What to prepare
Have a native reader check names and idioms. Keep each line short enough to sit beside the first language without a third line.
A worked example
Say your Reels are in English and the Audience view shows a large follower group in Brazil and another in Japan. Portuguese and Japanese are both in the 15 languages Storyboard18 lists, so either can go in Edits' second track. Pick the larger group first and add the other later.
If the larger group instead speaks a language outside the list, translate the cue sheet yourself and burn it, as the table says. Always have a native reader check names and idioms.
Sources
Related posts
More in Use cases
- Whole-song AI music video: a 3-minute video for $11.68
A full 3-minute music video on Sume: one Lyria 3.5 track, 18 ten-second Wan 3.0 shots at 480p and a 3-minute Timeline render for $11.68.
- Will my AI Reel show on Instagram for TV? 13+ ratings, shared screens
Instagram says TV Reels are for shared viewing and follow 13+ ratings. What that means for AI Reels, what the page leaves out, and a frame check before posting.
- Will TikTok remove my AI avatar brand account as fake?
TikTok removed 86 million fake accounts in Q1 2026 and bans impersonation. A labeled, invented avatar account is a different case, but TikTok makes no promise.
- Wine bottle reveal clip: 6 seconds of Wan 3.0 from a label photo
Animate a bottle photo into a 6-second reveal with wan-3.0 on Sume: 38 cents at 480p, 75 cents at 720p, $1.50 at 1080p. Why the label needs checking.
Written by Sume