Kling 3.0 Pro audio on vs off: price for 5, 10 and 15 second clips

Kling Video v3 Pro (kling-3) costs $0.14 per second with audio off and $0.21 per second with audio on when you call it through Sume. A 5-second clip is $0.70 or $1.05, and a 15-second clip, the longest it accepts, is $2.10 or $3.15.

Sume does not list a Kling 4.0 model as of 2026-10-08. The Kling entry in the Sume catalog is Kling Video v3 Pro, so any plan that names Kling 4.0 needs to use this model id or wait until the catalog changes. Check GET /v1/videos/models for the current list.

The vendor price

The fal model page for Kling Video v3 Pro text-to-video states that every generated second is charged $0.112 with audio off or $0.168 with audio on, and adds a higher rate when voice control is used (fal page, read 2026-10-08). The page describes the model as text-to-video with cinematic visuals, fluid motion, native audio and multi-shot support.

Sume carries the audio-off and audio-on rates and multiplies each by 1.25. The Sume catalog does not document the voice-control rate, so do not budget for it through Sume.

The envelope

The envelope is narrower than Seedance or Wan 3.0, which accept references and up to 30 seconds, and it is the only video model in the catalog limited to 16:9, 9:16 and 1:1.

• Duration: 4 to 15 seconds.
• Resolution: 720p or 1080p.
• Aspect ratio: 16:9, 9:16 or 1:1.
• Inputs: text-to-video, image-to-video and an end frame. The catalog constraint is no reference_*_urls, so reference images, videos and audio are not accepted.
• Audio: supported, and priced separately as above.

What the default costs

Sume's own test fixture for kling-3 with a 5-second text prompt reserves $1.05, which is 5 x $0.168 x 1.25. That means an estimate with no audio setting is priced as audio on. If you want the cheaper rate, say so explicitly in your request with generate_audio: false on /v1/videos, and check usage.cost on the first job to confirm.

curl -X POST https://api.sume.com/v1/videos \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kling-3","prompt":"A chef plates a dessert, close-up","resolution":"1080p","aspect_ratio":"16:9","duration":10,"generate_audio":false}'

Quick formula

Cost = seconds x $0.14 for audio off, or seconds x $0.21 for audio on. Both are the fal list of $0.112 and $0.168 times 1.25. A 4-second clip, the minimum, is $0.56 or $0.84.

Cost versus the alternatives

At 5 seconds with audio on, Kling v3 Pro is $1.05. The same length on Gemini Omni Flash 1.1 at 1080p, which always has audio, is $0.9375; on Wan 3.0 at 1080p it is $1.25. Kling with audio sits between Omni at 1080p ($0.1875 per second) and Wan 3.0 at 1080p ($0.25 per second), so the choice rests on the look of the output and on the envelope, not on the price alone.