Sound Scene00:28Futuristic Crisis Broadcast
Seed Audio 1.5 AI Audio Generator
Create expressive voices, conversations, music and complete sound scenes.
0/2048
60 Credits/min
Hear what a scene can become
From intimate voices to cinematic worlds. Explore the examples already available in our workspace. Seed Audio 1.0 examples
Sound Scene00:28
Voiceover00:31Li Mi's Memory
Conversation00:30Podcast Chat
Conversation00:30Crime-Thriller Call
6 min
Audio per generation
More room for a story to unfold.
6
Audio references
Bring more voices and sonic ideas.
30
Languages & variants
Tell the story to more audiences.
4
Input modes
Text, audio and video, together.
Beyond text to speech
A whole world of sound. One model.
Seed Audio 1.5 is ByteDance's audio generation model. It brings spoken dialogue, music, ambience and effects into one scene, with more ways to direct the performance and shape the final edit.
Start with words, guide a voice with references, or build sound around footage you already have. Longer generations and separate tracks open up a richer workflow for stories, podcasts and sound design.
Seed Audio 1.5 vs 1.0
More space. More references. More possibilities.
| Capability | Seed Audio 1.5 · Coming soon | Seed Audio 1.0 |
|---|---|---|
| Audio per generation | Up to 6 minutes | Up to 2 minutes |
| Audio references | Up to 6 | Up to 3 |
| Languages & regional variants | 30 languages & regional variants | 20+ languages |
| Input combinations | Text; text + audio; text + video; text + audio + video | Text; text + audio; text + image |
Seed Audio 1.5 capabilities
Direct the sound, down to the detail.
Write the moment, shape the performance and leave room for the edit.
Complete scenes, together
Generate dialogue, music, ambience and sound effects together, so the scene has a shared mood and direction.
Let the story breathe
Create a scene up to six minutes long. Give a conversation, narration or soundscape time to develop.
Six references to guide you
Bring up to six reference recordings into a generation to guide voices and the sound of your scene.
Stories across languages
Support for 30 languages and regional variants expands the possibilities for dialogue, narration and localization.
A performance with character
Guide reference voices with emotion, style and pace. Direct non-speech vocal sounds such as laughter, sighs and gasps.
Your timing. Your final mix.
Use timestamp instructions for dialogue, effects and music, then work with separate dialogue, ambience, effects and music tracks.
Four ways in
Start with what you have.
A script, a voice reference, a video or all three. Seed Audio 1.5 is designed to understand the material behind your story.
Text
Describe the voices, the setting and the sound you imagine.
Audio scene
Text + audio
Add reference recordings to guide the voice and sonic direction.
Audio scene
Text + video
Build dialogue and sound design around existing footage.
Audio scene
Text + audio + video
Combine voice direction with the visual context of your scene.
Audio scene
Sound that sees the scene
Give your footage a voice. And a world.
Build dubbing and a soundtrack around existing footage. Use the visual scene as context, then direct the dialogue, effects and music with timestamp instructions.
- 00:02
An alarm cuts through the engine hum.
- 00:06
A calm voice: “We still have time.”
- 00:12
The music builds as the ship departs.
Built for the next edit
One generation. Room to make it yours.
Separate dialogue, ambience, effects and music tracks give you more control after generation. Bring the layers into your editor to adjust the balance, refine a cue or reshape the music around a voice.
00:0000:1500:3000:45
Dialogue
Ambience
Effects
Music
From an idea to a sound scene
Turn your first idea into a complete sound scene, then refine the final mix.
- 01
Set the scene
Write your script and describe the atmosphere, speakers and sounds.
- 02
Bring your references
Add reference recordings, video or both to guide the voice and the world around it.
- 03
Direct the performance
Specify emotion, pace and timestamp cues for dialogue, effects and music.
- 04
Shape the final edit
Work with the separate tracks in your editor to finish the mix.
Made for your next story
More than a voiceover.
Explore how SeedAudio 1.5 fits into your creative process.
Audio stories & drama
Build character conversations with expressive performances, music and a sense of place.
Podcasts & narration
Explore longer spoken scenes, reference-guided voices and natural pacing.
Dubbing & localization
Use existing footage as context and explore dialogue across languages and regional variants.
Film & game sound design
Sketch layered soundscapes, cue effects to a moment and continue the edit with separate tracks.
FAQ
Seed Audio 1.5 FAQ
Capabilities, creative workflows and access on TopMaker.
Is Seed Audio 1.5 free to try?
TopMaker offers one free trial generation. After that, further generations use credits. Check the workspace for current pricing.
When will SeedAudio 1.5 be available on TopMaker?
The current editor and examples on this page use Seed Audio 1.0. Version 1.5 is not yet integrated into TopMaker, and its launch date has not been confirmed.
How does Seed Audio 1.5 compare with 1.0?
Compared with 1.0, the maximum generation length increases from two to six minutes, and the reference limit rises from three recordings to six. Version 1.5 also supports video inputs and separate tracks, with 30 languages and regional variants across four input combinations.
Can I control a voice’s emotion and speaking style?
Yes. Describe the emotion, speaking style and pace you want in your prompt, and use voice references to guide the performance. You can also request laughter, sighs and gasps. Results can vary, so you may need to refine your instructions to get the delivery you want.
Can I generate dialogue, music and sound effects together?
Yes. You can describe the dialogue, music, ambience and sound effects in a single prompt to create a complete sound scene. Reference recordings can help guide the voices and the setting.
Can I use voice references or video clips?
Yes. You can start with text alone, add reference recordings, include a video clip, or combine all three. Recordings guide the voices and sonic direction, while video provides context for dubbing and sound design.
Can I edit individual tracks after generation?
With version 1.5, you can work with separate dialogue, ambience, effects and music tracks in your editor. These layers are not available in the current 1.0 workspace.
Your next story starts with a sound.
Explore the workspace. Try a scene, bring a voice reference and shape your next audio story.
Create with Seed Audio 1.5TopMaker AI is an independent service and is not affiliated with or endorsed by ByteDance. TopMaker currently offers Seed Audio 1.0.