Seedance 2.0 Audio Generation: How Native Sound Works on Kensa (Complete Guide)

Seedance 2.0 Audio Generation: How Native Sound Works on Kensa

Seedance 2.0 generates audio together with the video: sound effects and ambient sound that match what happens on screen. On Kensa you control it with one switch, Generate Audio, which is on by default, and a clip with sound costs exactly the same number of credits as a silent one. This guide explains what the audio feature does and doesn't do, how to write prompts that get better sound, what it costs, and how it compares with the other models on Kensa, including Veo 3.1, which also produces audio.

Updated September 28, 2026: We rewrote this guide to match what Kensa actually offers. Earlier versions described features Kensa does not have and got Veo 3.1 wrong: every Veo 3.1 clip on Kensa includes audio.

Quick Answers

What Seedance 2.0 Audio Does

The sound is generated together with the video rather than picked from a library afterwards, so it follows what happens on screen. Typical results include:

• Sound effects tied to on-screen actions: footsteps, a door closing, a glass set down on a table, a car passing
• Ambient sound that fits the setting: rain, wind, waves, traffic, birdsong, the murmur of a café

You steer the audio with the words in your prompt. There is no separate audio prompt, no volume or mix control and no way to pick a specific track: the model decides the sound from your description and the picture. The result comes embedded in the video file you download.

What It Doesn't Do on Kensa

• No audio inputs. You can't upload music, a voice recording or sound samples to guide or replace the audio. Kensa accepts images only.
• No voice or script control. There is no field for dialogue or a voice to follow. If a clip needs specific spoken words, record a voiceover and add it in your editor.
• No separate audio track. You get one video file with the sound mixed in, not separate music and effects stems.

How to Turn On Audio

1. Open the Seedance 2.0 generator. Go to the Seedance 2.0 video generator. You can start from a prompt alone, add an image to animate, or add two images to fix the first and last frames.
2. Write your prompt with the sound in mind. Describe what the viewer should hear as well as what they should see. The tips below help.
3. Check Generate Audio. It sits in the generator's settings menu (the ⋯ button next to the other options) and is on by default. Leave it on for sound, or switch it off for a silent clip.
4. Pick length, resolution and aspect ratio. Seedance 2.0 offers 4 to 15 seconds, 480p or 720p, and six aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4 and 21:9). The credit cost shown before you generate is the same with audio on or off.
5. Generate, preview and download. Play the clip with sound before you download it. If the audio isn't right, adjust the sound descriptions in your prompt and generate again.

Learn more about the model itself on the Seedance 2.0 model page.

What It Costs

Seedance 2.0 is priced per second, and audio adds nothing. Regular prices on Kensa:

If a promotion is running, the generator shows the reduced price before you submit. If a generation fails, the credits held for it go back to your balance. See Pricing for plans and credit packs.

Audio Across the Models on Kensa

Seedance 2.0 vs Veo 3.1 for Sound

Both give you clips with sound for no extra credits, so the choice comes down to the rest of the clip:

• Veo 3.1 makes 8-second 720p clips, and every clip includes audio for a flat 13 credits. It is the simplest option when an 8-second clip is the right length. There is no switch to turn its audio off. See the Veo 3.1 model page.
• Seedance 2.0 lets you choose any length from 4 to 15 seconds, offers six aspect ratios including 21:9, can fix the first and last frames, and lets you switch audio off. It is priced per second, so it costs more per clip than Veo 3.1.

A practical workflow: test the idea and the sound with a short clip, then generate the final length once the prompt works.

Seedance 1.5 Pro

Seedance 1.5 Pro also has the Generate Audio switch at no extra cost, goes up to 12 seconds and offers 1080p output. Per second it costs less than Seedance 2.0, so it is the budget Seedance option when you don't need clips longer than 12 seconds or reference images.

Prompt Tips for Better Audio

Seedance 2.0 reads your prompt to decide what the clip should sound like, so describing the sound is the main lever you have.

Describe the Sounds You Want

The model adds plausible sound on its own, but naming the sounds gives it a much clearer target.

Weaker prompt: A chef cooking in a kitchen.

Stronger prompt: A chef tosses vegetables in a hot wok, oil crackling and popping, steam hissing as water hits the pan, utensils clinking and an extractor fan humming in the background.

Set the Sound Atmosphere

Words like quiet, bustling, echoing, muffled and crisp shape the ambient layer.

Example: A quiet library, pages turning softly, distant muffled footsteps on carpet, the faint hum of lights overhead.

Use Sound Words Sparingly

Words like whoosh, crackle, buzz and thud are clear cues. One or two per prompt is enough.

Example: A sports car accelerates with a deep rumble that builds into a roar, tires screech through the corner, then a whoosh of wind as it passes the camera.

Layer from Foreground to Background

Describe the sound the way a sound designer would build it: the main sound first, then the middle layer, then the background.

Example: Foreground: heels clicking on wet pavement. Mid-ground: light rain and distant traffic. Background: a church bell tolling far away.

Match the Sound to the Action

Keep the energy of the sound in line with the picture. Calm ambience under fast action, or loud effects over a still scene, feels wrong.

Know When to Add Audio Later

Turn Generate Audio off, or plan to replace the sound, when you need licensed brand music, a specific voiceover or exact dialogue. Kensa has no way to supply those to the model, so they belong in your editor. Switching audio off doesn't reduce the price.

Where Built-In Sound Helps

TikTok, Reels and Shorts

A clip that arrives with matching ambient sound and effects is ready for a first test post without a trip to a sound library. You can still layer trending audio or a voiceover on top in the platform's editor.

Product Clips

Small sounds make product footage feel real: a drink being poured, a lid clicking shut, a pan sizzling, a zip being pulled. Describe the action and its sound together.

B-Roll and Atmosphere

Establishing shots such as a city street at night, waves on a beach or a busy market are more convincing with their ambient sound, and you don't have to find a matching stock track.

Frequently Asked Questions

Does Seedance 2.0 audio cost extra credits?

No. A clip with Generate Audio on costs exactly the same as the same clip with it off.

Can I generate a Seedance 2.0 video without audio?

Yes. Switch off Generate Audio in the generator's settings and you get a silent clip.

Can I upload music, a voice recording or an audio reference?

No. Kensa accepts images as inputs (for image-to-video, first and last frames, and reference images), not audio or video files. Add your own music or voiceover in an editor after you download the clip.

Does Kensa offer lip-sync with Seedance 2.0?

No. Kensa doesn't offer lip-sync: there is no way to give a character a script or a voice to follow. For talking-head videos with exact dialogue, record the voiceover separately or use a dedicated avatar tool.

Does Veo 3.1 generate audio on Kensa?

Yes. Every Veo 3.1 clip on Kensa is 8 seconds at 720p and includes audio, for a flat 13 credits.

Does Kling 3 generate audio on Kensa?

Kling 3 has a sound option on Kensa, but the model is currently under maintenance and can't be used for new videos.

Can I use the videos commercially?

Commercial use of videos made on Kensa is included with the Pro and Ultimate plans and the Premium credit pack. See Pricing for details.

Getting Started

If you are new to Kensa, sign up with Google to get 15 free credits, valid for 3 days. That covers one Veo 3.1 clip with sound (13 credits); a regular-price Seedance 2.0 clip starts at 31 credits.

Start with a scene that has clear sound elements, such as a rainstorm, a busy café or a sizzling pan, generate it on the Seedance 2.0 video generator with Generate Audio on, and listen to how the sound follows the picture.

For everything else Seedance 2.0 can do, read the Seedance 2.0 complete guide. To animate photos cheaply when you'll add the sound yourself, see the Grok Image Video guide.