Audio Models

Choose the right audio model for your use case: text-to-audio, music, speech, sound effects, and video-to-music

ComfyUI Partner Nodes give you access to hosted audio generation models from providers such as ByteDance and Sonilo, so you can generate music, speech, and sound effects without a local GPU. This page helps you pick the right model for text-to-audio, music, or video scoring.

Browse audio models

Seed Audio 1.0

Text-to-audio and text-to-music, including instrumental and vocal tracks

Sonilo

Generate a full soundtrack that matches a video’s rhythm and content

What you can do with audio models

• Text-to-audio: Generate music or sound effects from a text description
• Video-to-music: Create a soundtrack that matches a video’s content and pacing
• Voice & speech: Generate speech or singing from text and reference audio

Choosing a model by task

Music & soundtracks

• Seed Audio 1.0 (ByteDance) — text-to-audio and text-to-music generation, including instrumental and vocal tracks
• Sonilo (Sonilo) — generate a full soundtrack from a video: it analyzes the video’s rhythm and content to produce a matching score

Sound effects & utility

• Seed Audio 1.0 (ByteDance) — also supports sound effects and audio-style transfer

All audio partners

Partner	Models
ByteDance	Seed Audio 1.0
Sonilo	Sonilo

See Pricing for per-model rates, and Concurrency Limits for usage caps.