Hunyuan Models

Video & Audio

1. hunyuan-video-foley

AI video foley and sound design that automatically adds footsteps, ambience, and effects while matching the timing and context of your footage.

Video Generation (2D)

1. hunyuan-video/i2v

Image-to-video generator that turns a single frame into a short, smooth clip while preserving composition and scene structure.

1. hunyuan-video/t2v

Text-to-video model for quickly generating stylized clips from natural-language prompts.

1. hunyuan-video-1.5/image-to-video

Upgraded image-to-video model with stronger temporal stability, richer detail, and more cinematic camera motion.

1. hunyuan-video-1.5/text-to-video

Version 1.5 text-to-video model offering better prompt adherence, cleaner structure, and higher overall visual quality.

Image & Avatar

1. hunyuan-image-2.1

Efficient general-purpose image generator that produces clean, realistic images from text prompts for a wide range of use cases.

1. hunyuan-avatar

Portrait-focused model for stylized avatars and character art, with vivid color, expressive faces, and consistent personal style.

1. hunyuan-image-3 & instruct

Advanced image model with stronger 3D awareness and scene composition, ideal for intricate diorama-style and environment-heavy visuals.

3D Models

1. hunyuan3d-v2-multi-view

3D multi-view generator that produces consistent renders of a character or object from multiple angles for downstream 3D work.

1. hunyuan3d/v2-base

Base 3D generator that creates solid, stylized shapes suitable for further sculpting, shading, or printing.

1. hunyuan3d/v2-mini

Lightweight 3D model specialized for compact, toy-like characters and small, stylized objects.

1. hunyuan3d/v2-multi-view

Multi-view variant of the v2 3D model that outputs coherent character views across several camera positions.

1. hunyuan3d/v2.1

Enhanced v2.1 3D generator with higher geometric fidelity and cleaner structure for polished stylized characters.