Does AI video audio cost extra? Kling adds 7 cents a second on Sume

Audio costs extra only on some models. Google's pricing page says all Veo 3.1 prices include video with audio by default, while Sume's Kling Video v3 Pro is $0.14 a second with audio off and $0.21 with audio on, a 50 percent surcharge of 7 cents a second. Sume's H3, H3 Max and Omni models always produce audio and have no toggle.

Audio on each side

Sume figures are the repository list price times 1.25.

What it adds up to

If a clip will carry a voiceover or a music bed added later, the generated audio is waste. Kling with audio off avoids paying for it, and Sume's separate tools can add the real sound track afterwards.

• A 10-second Kling clip is $1.40 silent and $2.10 with audio, so sound is $0.70.
• A 15-second clip, the maximum, is $2.10 silent and $3.15 with audio.
• On H3 and Omni you pay for audio inside the per-second rate, so you cannot save by omitting it.

Choosing

Ask whether the audio in the clip will survive to delivery. If a narration track replaces it, take a silent model setting. If the clip's own sound is part of the shot, such as dialogue or ambient noise, pay for it once at generation. For a timeline that mixes several clips, one consistent audio spine is usually better than the separate generated sounds of each clip.

Method

The Veo statement comes from the Gemini API pricing page, which also notes you are charged only if a video is generated successfully. The Sume figures are the repository list prices of $0.112 and $0.168 a second multiplied by 1.25. That gives $0.14 and $0.21, so the surcharge is $0.07 a second, which is 50 percent. The catalog marks Grok Imagine Video 1.5 as having no audio capability, and as image-to-video only.

If you cannot tell whether you need generated audio, run the same prompt both ways on Kling for a 5-second clip, which costs $0.70 silent and $1.05 with audio, and listen before you decide for a whole campaign.