← Back

See what the models can do.

Seed Audio 1.0

Let the scene be heard.

Explore dialogue, music and sound design in the audiovisual demos below. In Kivo, describe the scene, add audio references when needed, and generate an audio asset for your project.

  • Dialogue & atmosphere
  • Audio references
  • WAV / MP3
Sound for a scene

A video demo with generated audio. Turn on sound to hear the example.

These examples demonstrate the underlying models and were not made in Kivo. Results vary with inputs. Available controls in Kivo are described alongside each model.

Describe a scene you can hear

Seed Audio Create 1.0 can be used for more than isolated lines of speech. Describe the location, background texture, foreground effects and any dialogue together. For a game forest, specify distant birds, nearby leaves and the absence of music; for a conversation, separate the speakers and give each an emotion and speaking style.

Use timing and references with intent

You can describe a target length and the order in which sounds enter. Add an audio reference with @ when its voice or ambience matters, and explain its role. Kivo sends audio references as @ material labels for this model. Check the finished pacing: a timing request is creative guidance and should still be reviewed by listening.

Save the format you need, pay for actual duration

Kivo offers WAV and MP3 output for the current audio workflow. Credits are calculated from the actual generated duration at the configured per-second rate and rounded to a whole number after calculation. The studio shows the accepted maximum charge before submission; the duration is not fixed by a mandatory slider.

Audio Create 1.5 is coming soon

Audio Create 1.5 is planned for Kivo. It is not an active model choice yet. Continue building reusable voice and ambience references with 1.0; supported inputs, formats and pricing for 1.5 will be shown when its integration is ready. No additional capabilities or release date are promised here.

From the examples to your project

Listen for the relationship between a voice, its delivery and the atmosphere around it in the examples above. For your own scene, describe who speaks, the mood, the words and the desired background sounds. Use time ranges when a line or effect needs a particular place in the scene.

Kivo accepts audio references and produces WAV or MP3 audio assets. The videos above are listening demos, not video outputs from Kivo’s audio tool. Review pronunciation, pacing and sound balance before using an output in a final edit.

Start creating