by Tencent
Consumer-oriented successor to HunyuanVideo 13B from Tencent. 8.3B parameter 3D DiT supporting both text-to-video and image-to-video (T2V + I2V). Step-distilled variant runs 480p at ~75s on RTX 4090; minimum ~14GB VRAM with offload in FP16.
Your hardware
Detecting...
Measured quality metrics for HunyuanVideo 1.5 outputs.
How often humans prefer this model's output (0-100%)
Visual quality and composition rating (5-9 scale)
VRAM estimates at FP16 and FP8 precision. FP8 uses ~40% less memory with minimal quality loss. Grade shows how well each GPU handles the generation workload.
| Scenario | VRAM | RTX 4090 24GB | RTX 3060 12GB | RTX 4060 8GB | MacBook Pro M4 Pro 24GB |
|---|---|---|---|---|---|
| 512×512 · 25 frames | 37.1 GB | F | F | F | F |
| 768×512 · 25 frames | 39.2 GB | F | F | F | F |
| 768×512 · 100 frames | 45.5 GB | F | F | F | F |
| 1280×720 · 25 frames | 47.6 GB | F | F | F | F |
| Scenario | VRAM | RTX 4090 24GB | RTX 3060 12GB | RTX 4060 8GB | MacBook Pro M4 Pro 24GB |
|---|---|---|---|---|---|
| 512×512 · 25 frames | 21.0 GB | A | F | F | D |
| 768×512 · 25 frames | 23.1 GB | B | F | F | D |
| 768×512 · 100 frames | 29.4 GB | D | F | F | F |
| 1280×720 · 25 frames | 31.6 GB | D | F | F | F |
Turbo / LCM distillation
Use distilled scheduler at 4-8 steps for faster iteration
from diffusers import HunyuanVideoPipeline
import torch
pipe = HunyuanVideoPipeline.from_pretrained(
"tencent/HunyuanVideo-1.5",
torch_dtype=torch.float16
)
pipe.to("cuda")
frames = pipe(
prompt="your prompt here",
num_inference_steps=50,
guidance_scale=6.0,
num_frames=129,
).frames[0]
# Save frames or export as videoGet started
Setup instructions for running HunyuanVideo 1.5 locally
1. Download the model
Get the checkpoint from HuggingFace
2. Place in:
ComfyUI/models/checkpoints/3. Launch ComfyUI
python main.pyVRAM allocation for 25 frames at 768×512 on RTX 4090 24GB
25 frames at 768×512, 30 steps, FP16.
Download HunyuanVideo 1.5 in different precisions. Lower precision = less VRAM but slight quality loss.
| Format | Precision | Size | Provider | |
|---|---|---|---|---|
| safetensorsRecommended | FP16 | 16.0 GB | official | Download |
Early LoRA ecosystem following on from the HunyuanVideo community.
Frequently asked questions
HunyuanVideo 1.5 (8.3B parameters) requires approximately 39.2 GB of VRAM at FP16 precision for generating 25 frames at 768×512. Video generation typically requires more VRAM than image generation due to temporal attention layers.
HunyuanVideo 1.5 exceeds the RTX 4090's 24 GB VRAM at FP16 for video generation. Consider reducing resolution, frame count, or using a GPU with more VRAM.
On a reference GPU (RTX 4090 24GB), HunyuanVideo 1.5 generates a 25-frame video at 768×512 in approximately ~2m 33s at FP16 with 30 inference steps. Faster GPUs with higher memory bandwidth will reduce generation time.
HunyuanVideo 1.5 supports up to 1280×720 resolution and 129 frames per generation at 24 FPS. Higher resolutions and frame counts require proportionally more VRAM.
See also