# Seedance 2 Reference

> Video AI Model by ByteDance — available on Layer

Generate videos guided by reference images, videos, and audio with precise style and character control.

Seedance 2 Reference is the reference-guided video generation mode of ByteDance's Seedance 2 model, designed for creators who need fine-grained control over style, character consistency, and audio synchronization. It accepts up to 9 reference images, 3 video clips, and 3 audio files alongside a text prompt, enabling multi-modal guidance that goes beyond traditional text-to-video workflows. This makes it an ideal tool for maintaining visual identity across scenes, adapting motion from existing footage, and producing branded video content with consistent characters.

## Specifications

| Property | Value |
|----------|-------|
| Provider | ByteDance |
| Category | Video |
| Price | 7 / megapixel second per megapixel_per_second |
| Aspect Ratios | 1:1, 4:3, 3:4, 16:9, 9:16, 21:9 |
| Durations | 4s, 5s, 6s, 7s, 8s, 9s, 10s, 11s, 12s, 13s, 14s, 15s |

## Capabilities

- Audio Generation

---

[Try Seedance 2 Reference](https://next.app.layer.ai/me/session/new?baseModelId=SEEDANCE_V2_REFERENCE) | [All Video Models](https://layer.ai/models/video) | [All Models](https://layer.ai/models)