Video Generation, Finally With Sound: HunyuanVideo-Foley Now on WaveSpeedAI

For years, AI video generation has been a silent movie. We’ve watched stunning AI-created visuals unfold in eerie quiet, waiting for technology to find its voice.

Today, that wait is over. HunyuanVideo-Foley, Tencent Hunyuan’s end-to-end video sound effect generation model, is now available on WaveSpeedAI. For the first time, creators can generate synchronized video and cinematic-quality audio—powered by the world’s fastest multimodal inference acceleration platform.

You only need to upload a silent video and a simple text prompt, and the result will be a realistic, sound-enabled video.

Solving the Three Biggest Challenges in AI Audio

Existing audio generation technologies face several difficulties, such as limited generalization, inconsistent semantic alignment, and poor audio quality.

HunyuanVideo-Foley has solved each of these issues one by one.

The highlights are:
🎭 Multi-Scene Synchronization – High-quality audio aligned with complex visual scenes
🧠 Multi-modal Balance – Perfect harmony between visual and textual cues
🎵 48kHz Hi-Fi Output – Professional-grade clarity, free from noise and artifacts
🏆 SOTA Performance – Leading across all benchmarks in fidelity, synchronization, and semantic alignment

From Short Clips to Blockbusters

In fact, HunyuanVideo-Foley is groundbreaking. It can be applied to a wide range of scenarios. For example, by uploading a silent video, it can instantly generate a soothing cooking ASMR clip.

Silent Video Prompts: A relaxing ASMR video of hands slicing fresh kiwi fruit on a wooden cutting board. Close-up macro view, crisp knife sounds, juicy textures, soft natural lighting, clean minimal background. Emphasis on tactile ASMR sounds: slicing, peeling, gentle tapping.
Text Prompts: Generate realistic cutting sounds of a kiwi.

At the same time, At the same time, dubbing for movie-level scenes is equally effortless.

We can also add all the sounds that animated works are supposed to have.

There are many more applications for HunyuanVideo-Foley—you can explore them at your own pace.

Start Creating Today

The era of silent AI video is officially over.
Video production technology is advancing rapidly, and creators must embrace these tools to produce higher-quality content.
Try HunyuanVideo-Foley Today：https://wavespeed.ai/models/wavespeed-ai/hunyuan-video-foley

Besides, you can connect with us on social media below.
Discord: http://discord.com/invite/yHa3J4Vcd
X (Twitter): http://x.com/wavespeed_ai
Open Source Projects: http://github.com/chengzeyi

© 2025 WaveSpeedAI. All rights reserved.