AI models on Craze
Every video, image, voice, music and sound model Craze can direct, with what each one makes and what it costs.
Every video, image, voice, music and sound model Craze can direct, with what each one makes and what it costs.
Starts at$0.011 per second of source video
Tencent's model that generates professional sound effects directly from video. Describe your scene, provide a clip, and Craze generates synchronized audio synced to every movement.
Foley work usually means watching footage frame by frame and manually timing every footstep, door slam, and rustling fabric.
The model generates sound at 48 kHz, the standard for professional film and television.
You don't have to rely on the visual content alone.
Hunyuan Video Foley was trained on 100,000 hours of multimodal video and audio, so it handles a wide range of scenes and sound types, from dialogue and music cues to natural ambience and mechanical sounds.
Craze creatively jams, researches and plans with you.
Craze generates anything imaginable. Video, audio, music, motion graphics and more.
Craze watches your footage and builds a final cut. With a real editor if you need it.
Every model here runs on every Craze plan, including Free, and Craze can use a different one for each shot.
| Hunyuan Video FoleyTencent | ElevenLabs Sound Effects V2ElevenLabs | |
|---|---|---|
| Makes | Text + Video to audio | Text to audio |
| Length | — | Up to 22s per clip |
| From | $0.011 per second of source video | $0.003 per second of audio |
| Task | Inputs | Resolution | Aspect ratios | Sound |
|---|---|---|---|---|
| Sound for video | Video | Follows the source | Follows the input | None |
$0/ month
$19/ month
$59/ month
$199/ month
Hunyuan Video Foley is a video-to-audio model from Tencent that generates synchronized sound effects and ambient audio directly from video footage. Instead of layering audio separately in post, you describe what you want to hear, provide your video, and the model creates sound that moves with the picture.
Sound is often the last thing creators tackle, and it's the first thing audiences notice when it's missing. Hunyuan Video Foley changes that by embedding audio generation into the video workflow, reading both motion and intent.
Foley work usually means watching footage frame by frame and manually timing every footstep, door slam, and rustling fabric. Hunyuan Video Foley watches the video with you, understands what's moving and where, and generates audio that lands with the action. A character's footsteps sync to their stride. A hand reaches for an object and the model times the contact sound.
The model generates sound at 48 kHz, the standard for professional film and television. It uses a custom audio VAE trained to minimize noise and artifacts, so the generated sound is clean enough to sit alongside other professional audio elements.
You don't have to rely on the visual content alone. Describe the atmosphere, the emotional tone, or specific details you want to hear, and Hunyuan Video Foley balances your text prompt with what it sees in the video. A wide shot of a city generates ambient traffic and distant voices. The same shot with a prompt like "quiet morning" pulls back the energy and fills space with birdsong instead.
Hunyuan Video Foley was trained on 100,000 hours of multimodal video and audio, so it handles a wide range of scenes and sound types, from dialogue and music cues to natural ambience and mechanical sounds.
Craze uses Hunyuan Video Foley by default when you need audio synced to video. The workflow is straightforward.
If you want to name another audio model, you can. Layer with music from ElevenLabs Music or add dialogue with ElevenLabs v4. But for sound that moves with picture, Hunyuan Video Foley handles the timing automatically.
Sound generation is billed $0.011 per second of source video, calculated on the length of your source video. The model generates continuously, so a 60-second clip costs twice as much as 30 seconds. Use the cost calculator on this page to estimate before you generate.
Audio post-production often comes last, after picture lock. Hunyuan Video Foley lets you generate realistic sound without syncing manually. Sit a speaker in a room, light it well, and generate the audio that brings the scene to life.
Animated sequences and game cutscenes are often made without sound. Hunyuan Video Foley reads the motion and generates synchronized foley and ambience, making animation feel more grounded and real.
Short product videos often feel hollow because sound is an afterthought. Generate sound that moves with product motion: a lid clicking shut, liquid pouring, a button press. The video feels more polished and intentional.
Both models generate audio from text prompts, but they approach the problem differently. Hunyuan Video Foley reads video motion and generates sound locked to the picture. ElevenLabs Sound Effects V2 generates isolated sound effects from text alone, without video context. Use Hunyuan Video Foley when you need timing precision and sync to motion. Use Sound Effects V2 when you want isolated, layerable elements to place on the timeline yourself.
Describe what you want the way you would brief an editor, and Craze writes the Hunyuan Video Foley prompts for you. Every Craze plan, including Free, runs Hunyuan Video Foley and every other media model, and Craze can use a different one for each job. Each sound effect lands on the timeline under your cut, and exports with the video or on its own as MP3. Generations come out of your plan's monthly credits, from $0.011 per second of source video, with top-ups when you need more.
Video editors adding production sound: Audio post-production often comes last, after picture lock. Game developers and animators: Animated sequences and game cutscenes are often made without sound. UGC and product demo creators: Short product videos often feel hollow because sound is an afterthought.
Hunyuan Video Foley is a video-to-audio model from Tencent that generates sound effects and ambient audio directly from video clips. It reads visual motion and context to create synchronized, realistic audio.
Craze uses Hunyuan Video Foley by default when you want sound synced to existing video. Upload your clip and describe what audio you want, and the model generates sound that matches the visual timing and context.
Hunyuan Video Foley creates realistic sound effects, ambient audio, and foley sounds that match video motion and context. It handles everything from footsteps and object interactions to environmental sounds like wind or rain.
On Craze, you can upload video clips up to 5 minutes long. The model generates broadcast-quality audio at 48 kHz.
Hunyuan Video Foley is billed $0.011 per second of source video, based on the length of your source video. Use the cost calculator on this page for exact estimates.
Hunyuan Video Foley is available on every Craze plan, including Free.
Tencent open-sourced Hunyuan Video Foley on August 28, 2025, with inference code and model weights available to the community.
Hunyuan Video Foley is trained on 100,000 hours of multimodal data and generates sound by reading visual motion and text prompts together. It produces broadcast-ready audio at 48 kHz with minimal noise.
