What sound adds to the price of a generated second
Five channels here show what sound does to the price of a second, and the answers run from nothing to double. Audio adds roughly a third on one platform's own model, doubles one resold model, costs nothing extra on another, and on one channel speech comes inside the tier rate with an episode score at no charge. As of 2026-10-25.
| Channel | Without sound against with it | What the sound includes |
|---|---|---|
| Luma | Veo 3.1 at 140 or 280 credits a second; Kling 3.0 at 30 or 40 at 720p, the same at 4K | Native audio of each resold model |
| PixVerse | 9 or 12 credits a second at 720p; 18 or 23 at 1080p | Audio generated with the clip |
| Runway | Veo 3.1 at 20 or 40 credits a second | Native audio of the resold model |
| SceneMixer | One rate per tier; speech and ambience are in it | Each line in the character's voice, plus a free episode score |
| Wan | Switching audio on or off does not change the price | An audio track on by default |
Inclusion rule. Channels whose price list shows a rate for sound, or states that sound does not change the rate. A channel that prices video without mentioning audio is not given a row. Order. Alphabetical by channel.
1The same model can cost double with its voice on
Resold models carry their maker's audio pricing with them. One model doubles from 140 to 280 credits a second on one channel, and from 20 to 40 on another, which says the doubling belongs to the model rather than to either shop. Another model adds a third at 720p, about a fifth at 1080p and nothing at 4K, where the silent and voiced rates are the same.
For dialogue-heavy drama that line is the budget. A spoken scene generated without sound is not a cheaper version of the same deliverable; it is a scene that still needs a voice from somewhere.
2Included, or switched off for nothing
Two rows go the other way. One model's documentation says switching audio on or off does not affect the price, so silence saves nothing. One drama tool prices each tier once and states that a segment comes out with its sound in it: every line in that character's voice, the lips moving with the words, and the ambience of the place.
Where sound is in the rate, the comparison against a silent-plus-voiced price has to add the voiced surcharge back to the other side, or the cheaper-looking row is not cheaper.
3Music is a separate line on every list but one
Native audio on a shot is speech and effects, and whatever music a prompt happens to produce. A score that runs across an episode is a different product. One channel states that its episode score is composed and mixed in under the voices and costs nothing extra; elsewhere, music is either a separate model with its own price or left to the edit.
- What audio adds per secondVeo 3.1 is listed at 140 credits a second without audio and 280 with it, at 720p and at 1080p
- What audio adds per secondKling 3.0 is listed at 30 credits a second without audio and 40 with it at 720p, 40 and 49 at 1080p, and 147 either way at 4K
- What audio adds per secondV6 costs 7, 9, 12 or 23 credits a second with audio at 360p, 540p, 720p or 1080p, against 5, 7, 9 or 18 without
- What audio adds per secondVeo 3.1 on the developer API is 40 credits a second with audio and 20 without
- What audio adds per secondEnabling or disabling audio does not affect pricing
- What the per-second price includesEach segment comes out with its sound in it: each line in the character's voice with lips moving to the words, and the ambience of the place
- Music for a whole episodeThe episode's background score is included with the videos and costs nothing extra
- Wan 3.0 480P10 credits per second, about $0.06the default tier
4Sources
Each figure is read from the vendor page linked in the ledger, checked 2026-10-25. How a figure qualifies is on method; what a rate leaves out is on limits and unspent credits and parallel generations. Nearby: Price of a reference, Remaking a clip, Waiting time.