Reels on the Instagram TV app: caption for a living room
Reels on a TV autoplay with sound, but a screen across the room needs bigger text. Raise font_size_ratio in a Sume caption job and test one frame first.

Make caption text larger than you would for a phone: raise typography.font_size_ratio in your Sume caption job, keep lines short, and check a single frame on a real TV before you batch. Instagram for TV autoplays Reels with full sound, but people watch it from a sofa, not at arm's length.
Instagram's own announcement says Reels on TV are organised into interest channels, play automatically, and support up to five accounts on one device; it also says the app was tested on Amazon Fire TV in the US and has since expanded to Google TV devices in the US (Instagram, read 2026-10-07). It does not publish caption rules for TV.
What the announcement does and does not tell you
There is no TV-specific Reel format in the page we read, so the same vertical file plays. What changes is how the viewer reads it.
| Item | What the page says |
|---|---|
| Devices | Amazon Fire TV (test), Google TV devices in the US |
| Organisation | Interest-based channels with autoplay |
| Accounts | Up to five per device |
| Audio | Plays with full sound |
| TV caption or ratio rules | Not stated |
Settings to change in a Sume caption job
Video captions take a design block per request. The API accepts typography.font_size_ratio from 0.01 to 0.4, typography.safe_width_ratio from 0.3 to 1, phrasing.max_words from 1 to 12 and placement.anchor_ratio from 0.05 to 0.95. Out-of-range values return a 400 before any render.
A starting point for a TV-first cut is fewer words per card and a middle anchor: max_words 3, anchor_ratio 0.5, and a font size ratio a notch above what you use for a phone Reel. Raise it in steps, because the safe width stops a long word from running past the frame edge.
curl -X POST https://api.sume.com/v1/video-captions \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: tv-caption-test-01" \
-d '{
"video_url": "https://media.sume.com/artifacts/example/clean.mp4",
"style": "black-outline",
"design": {
"phrasing": { "max_words": 3 },
"placement": { "anchor_ratio": 0.5 }
}
}'A one-frame test
Pull a still at a busy moment with video frames (at: [4.5]), cast or open it on the TV, and look from where people sit. If you can read it, the caption is large enough. A restyle with source_caption_id runs no second speech-to-text, though each caption job is $0.20 for a clip of up to 60 seconds.
Do not forget the phone
A file with larger captions still has to work on a phone. A good compromise is one step up from your usual size, not a jump to the largest. Check both: the TV for legibility and the phone for crowding. If the two disagree, favour the phone, because most views still happen there. The announcement describes TV as an extra place for Reels, not a replacement for the app.
Keep a version for each
If you care about TV, keep two caption looks for the same cut. Using source_caption_id you can restyle without a second speech-to-text run, so the second look costs a render, not a transcript. Name the files phone and tv, and post the one you use most.
Sources
Related posts
More in Media tools
- Reels picture-in-picture test: check the four corners first
Instagram is reported to be testing picture-in-picture. Crop each corner of your 1080x1920 Reel with video-filter, free to check, $0.02 to encode.
- Mute a clip or extract its audio? Trim audio drop vs audio detach
Video trim with audio drop returns a silent MP4; audio detach returns a separate wav or mp3 and leaves the video alone. Which to use, and the warnings.
- Repurposing TikTok and Reels to Shorts: restyle captions once, $0.20
Reposting a clip to YouTube Shorts? Keep a clean master and burn captions with Sume: $0.20 a job for up to 60 seconds, and restyles skip a second transcription.
- Move burned-in captions up or down: restyle with source_caption_id
Change caption placement with design.placement.anchor_ratio and restyle with source_caption_id: the same word timings, no second speech-to-text run.
Written by Sume