One music bed, three TikTok ratios: three renders for $0.425
A 9:16, 16:9 and 1:1 TikTok ad can share one $0.125 bed. Each Timeline render at 1080x1920, 1920x1080 and 1080x1080 is $0.10 up to a minute, so $0.425 in all.

One generated bed can sit under all three TikTok ad ratios. Render the same cut three times on Timeline 1.0 at 1080x1920 (9:16), 1920x1080 (16:9) and 1080x1080 (1:1). Each render of 60 seconds or less is $0.10, and the bed is $0.125, so the sound-and-assembly cost is $0.425 for the set, or about $0.142 per ratio.
TikTok's specification for non-Spark ads (read 2026-10-08) sets the minimum sizes: 540x960 for 9:16, 960x540 for 16:9 and 640x640 for 1:1, with up to 10 minutes of video and a file of at most 500 MB.
Sizes and prices
Timeline accepts output.width and output.height as even integers from 256 to 2160, and its default output is 1080x1920. All three sizes are above TikTok's minimums. The footage must exist in a shape that fits each frame, and video[].fit chooses between cover, contain, stretch and blur for clips that do not.
| Ratio | TikTok minimum | Timeline output | Render (<= 60 s) |
|---|---|---|---|
| 9:16 | 540x960 | 1080x1920 (default) | $0.10 |
| 16:9 | 960x540 | 1920x1080 | $0.10 |
| 1:1 | 640x640 | 1080x1080 | $0.10 |
| Bed (once) | $0.125 | ||
| Total | $0.425 |
What stays the same, what changes
The spine and the soundtrack are identical across the three requests, so the music is the same length and has the same ducking. What changes is the picture: a vertical clip dropped into a 16:9 frame with fit: contain shows bars, while blur fills them. If you want true native framing for each ratio, the footage has to be generated per ratio, which is a footage cost the table does not include.
Gemini Omni Flash 1.1 makes 16:9 or 9:16 only, so a 1:1 version needs a crop or cover from one of those, or another model. Check the ratios of your video model in the catalog before you plan a three-ratio set.
When it is not worth a bed
A bed under three ratios saves two generations: three separate beds would be $0.375, not $0.125. But if the three versions are meant to sound different for different audiences, pay for three. The extra $0.25 is small against the cost of the footage.
Length is the other limit. A render is billed per started output minute, so a 61-second cut bills two minutes and the three-ratio set rises from $0.425 to $0.725. If your ad is 62 seconds because of a final card, trimming to 60 saves $0.30 on this set alone. Trim with the Timeline spine rather than a separate job, since the render already applies the cut.
The cheap check is to render the 9:16 first and listen. A re-roll costs $0.10; discovering a weak bed after all three are done costs $0.30.
Name each output after its ratio and keep the three Idempotency-Keys distinct (for example spring-9x16, spring-16x9, spring-1x1). Reusing one key across different ratios risks the second request being treated as a retry of the first, so give each ratio its own.
Sources
Related posts
More in Use cases
- One captioned 15-second clip for Bluesky, Threads and WhatsApp status
Make one 15-second muted-friendly clip with burned-in captions and post the same file on Bluesky, Threads and a WhatsApp status. Model options and cost.
- One product photo, five ad shapes: Flux 2 Pro $0.1875, Qwen $0.125
Five shapes (1:1, 4:5, 9:16, 16:9, 21:9) from one photo: $0.1875 on Flux 2 Pro, $0.125 on Qwen Image, $0.50 on Nano Banana 2.1. Seedream 5 Lite lacks 21:9.
- Open-house reel from 12 listing photos: one Timeline render, $0.10
12 listing photos at 3 seconds each make a 36-second vertical reel in one Sume Timeline render: one voice spine, fades, blur fit, ceil(36/60) = 1 minute, $0.10.
- A real person's photo avatar reading a script: YouTube disclosure
A Sume avatar made from a real person's photo and given a script can show them saying words they never said, which YouTube says needs a disclosure label.
Written by Sume