- Multi-modal 4-input system (image, video, audio, text) with up to 12 files per generation
- Universal @ reference syntax to assign roles to assets
- Built-in dual-channel stereo audio with lip-sync and music beat sync
- Camera replication and action replication from reference videos
- Video editing and extension (character swaps, scene insertion, smooth continuation)
- One-take continuity for spatial and temporal consistency
- Unified credits and subscription plans with daily free credits