Hunyuan Video Generator
Create AI videos with Tencent Hunyuan Video — the open-source video generation model with high-quality motion, strong prompt adherence, and FramePack optimization. Text-to-video and image-to-video up to 1080p resolution.
framepack makes Hunyuan practical on consumer GPUs by reducing VRAM pressure during inference. With framepack optimizations, creators can test more ideas locally,
Hunyuan Video Features
Fully Open Source
Hunyuan Video is fully open-source under a permissive license, enabling researchers, developers, and creators to run, modify, and build upon the model. Community-driven forks like FramePack have pushed local inference to consumer GPUs with as little as 6GB VRAM.
High-Quality Motion
Generates videos with remarkably smooth and physically plausible motion. Human movement, fluid dynamics, and camera trajectories feel natural and weighted — a key strength that sets Hunyuan apart from many open-source alternatives.
FramePack Optimization
The community-developed FramePack technique restructures input frames to dramatically reduce VRAM usage during inference. Run Hunyuan Video locally on a consumer GPU — generate videos without expensive cloud compute or enterprise hardware.
Strong Prompt Adherence
Hunyuan Video follows text prompts with high fidelity, accurately rendering described scenes, actions, and visual styles. Supports detailed multi-clause prompts for precise creative control over subject, motion, lighting, and camera.
Chinese Language Support
As a Tencent model, Hunyuan Video has native Chinese language understanding — it handles Chinese prompts with the same fidelity as English. Ideal for creators producing content for Chinese-speaking audiences or working with Chinese cultural concepts.
Fully Open Source
Hunyuan Video is fully open-source under a permissive license, enabling researchers, developers, and creators to run, modify, and build upon the model. Community-driven forks like FramePack have pushed local inference to consumer GPUs with as little as 6GB VRAM.
High-Quality Motion
Generates videos with remarkably smooth and physically plausible motion. Human movement, fluid dynamics, and camera trajectories feel natural and weighted — a key strength that sets Hunyuan apart from many open-source alternatives.
FramePack Optimization
The community-developed FramePack technique restructures input frames to dramatically reduce VRAM usage during inference. Run Hunyuan Video locally on a consumer GPU — generate videos without expensive cloud compute or enterprise hardware.
Strong Prompt Adherence
Hunyuan Video follows text prompts with high fidelity, accurately rendering described scenes, actions, and visual styles. Supports detailed multi-clause prompts for precise creative control over subject, motion, lighting, and camera.
Chinese Language Support
As a Tencent model, Hunyuan Video has native Chinese language understanding — it handles Chinese prompts with the same fidelity as English. Ideal for creators producing content for Chinese-speaking audiences or working with Chinese cultural concepts.
Fully Open Source
Hunyuan Video is fully open-source under a permissive license, enabling researchers, developers, and creators to run, modify, and build upon the model. Community-driven forks like FramePack have pushed local inference to consumer GPUs with as little as 6GB VRAM.
High-Quality Motion
Generates videos with remarkably smooth and physically plausible motion. Human movement, fluid dynamics, and camera trajectories feel natural and weighted — a key strength that sets Hunyuan apart from many open-source alternatives.
FramePack Optimization
The community-developed FramePack technique restructures input frames to dramatically reduce VRAM usage during inference. Run Hunyuan Video locally on a consumer GPU — generate videos without expensive cloud compute or enterprise hardware.
Strong Prompt Adherence
Hunyuan Video follows text prompts with high fidelity, accurately rendering described scenes, actions, and visual styles. Supports detailed multi-clause prompts for precise creative control over subject, motion, lighting, and camera.
Chinese Language Support
As a Tencent model, Hunyuan Video has native Chinese language understanding — it handles Chinese prompts with the same fidelity as English. Ideal for creators producing content for Chinese-speaking audiences or working with Chinese cultural concepts.
Start Creating with Hunyuan Video
Generate AI videos with Tencent's open-source Hunyuan Video model. High-quality motion, strong prompt adherence, and FramePack optimization — powerful video generation for everyone.
