What one localized video costs on Sume: captions, voiceover, dub
Per language, a translated caption job is $0.20, a voiceover on a silent clip up to $0.13, a 60-second dub up to $0.17 and a nine-minute lesson dub up to $1.37.
Per extra language, a translated caption job costs $0.20 on Sume, a voiceover on a silent 20-second clip up to $0.12, a 60-second dub up to $0.17, and a nine-minute lesson dub up to $1.37. Translation itself runs on Bilibili's listed free Index-Translate API, so it adds no Sume line.
These come from the same step rates, with the assumptions stated in the table: 15 characters per second for a voiceover and 850 characters per minute for a dub. Rates are from the pricing page, Video captions docs and Timeline docs, read 2026-10-05.
The price ladder, per language
Use it to pick the cheapest method that does the job.
| Method | Input | Steps billed | Cost, up to |
|---|---|---|---|
| Caption on a silent clip | Clip of up to 60 s | One caption job | $0.20 |
| Caption on a clip with speech | Clip of up to 60 s, 1 min of speech | Detach, STT, caption job | $0.22 |
| Voiceover on a silent clip | 20 s, 300 characters | TTS, concat, 1-min render | $0.124 |
| Dub of a 60-second video | 900 characters | Detach, STT, TTS, concat, render | $0.173 |
| Dub of a 9-minute lesson | 7,650 characters | Same, 9-minute render | $1.373 |
How to read it
- Under a minute, a dub can cost less than a caption job, because the render is $0.10 and speech is cents.
- Past a few minutes, the render at $0.10 a minute dominates.
- Run
POST /v1/timeline-1.0/plan, which is unbilled, to see the billable minutes before a render. - Each line is a ceiling; render and concat are captured at their own compute.
What it leaves out
It leaves out any video generation, avatar presenter or review time, and it assumes a clip already on media.sume.com. Billing for those steps is on the pricing page.
Sources
Related posts
More in Pricing
- One prompt, three ratios: a 1:1, 16:9 and 9:16 set priced by model
All 19 Sume image models list 1:1, 16:9 and 9:16. A three-image set costs $0.075 on Qwen Image, $0.137 on GPT Image 2.5 high, $0.56 on Nano Banana Pro.
- 15-second AI video via API: price on nine Sume model settings
A 15-second AI video via API runs from $0.19 on Grok Imagine to $21.32 on Seedance 2.5 at 1080p. One job each, priced from Sume's per-second rates.
- A 150 ms TTS claim vs Sume's async TTS: which jobs need which
MAI-Voice-2.1-Flash claims 150 ms end to end. Sume TTS is an async job at $0.0475 per 1,000 characters. Where each fits, and 1,000 chat turns at 200 characters.
- A 24-door advent calendar of AI clips: cost per door and per calendar
One card, one 5-second Wan 3.0 clip and one 200-character voice line per door: $8.95 for a light calendar, $17.29 mid-tier and $36.97 premium on Sume.
Written by Sume