FreeAn open 13.6B video model for text-to-video, image-to-video, and video continuation, designed for coherent longer clips.
Video & Animation
All tools
View more Free 
Free tier
Access
Listed as Tencent publishes the HunyuanVideo-Foley code and model resources openly, so the software and weights can be inspected and run for free under the project terms. The official setup needs Linux, CUDA, and a GPU with roughly 8 to 20 GB of VRAM depending on the model and offload mode. It should be described as a free open research workflow, not as an unlimited hosted audio generator.
0
Audio & Voice Checked Aug 6, 2026
HunyuanVideo-Foley
Generate scene-aware foley and sound effects that follow the action in a video.
Best for: Video editors, sound designers, and developers building an open post-production workflow.
What you can do with it
- Add footsteps or impacts to a silent clip
- Create environment sounds for a scene
- Prototype audio for a short film
- Test audio-visual synchronization
- Run a local foley generation demo
Good to know
- Official Tencent open-source project
- Scene-aware sound-effect generation
- 48 kHz audio support
- Local Gradio workflow
Limitations
- Requires local GPU setup
- Not a music or voice generator
- Hardware needs vary by model
Other official links
Keep exploring
Similar tools
FreeTransfer a performance to a character image or replace a person in a video with an open Wan model.
Video & Animation
Free tierTurn a prompt or still image into a short video with Alibaba's open Wan model family.
Video & Animation