HunyuanVideo 1.5 is Tencent's open video model. ComfyUI's own workflow downloads 29.0 GB and needs 19.2 GB on the card at its largest stage, the diffusion stage. The smallest card listed here that loads that set with room to spare is the RTX 4090. With a smaller file, the FP8 scaled version, it loads on an RTX 4070.
The licence does not apply in the European Union, the United Kingdom and South Korea. If you are there, this licence does not cover you.
What HunyuanVideo 1.5 is
Tencent's 8.3-billion-parameter video model from November 2025: 480p and 720p, 121 frames by default, with a separate super-resolution model that takes the result to 1080p. Small for a video model, with a licence that excludes the EU and the UK.
8.3B parameters Text encoder: Qwen2.5-VL 7B Tencent Hunyuan Community License November 2025
Before you download
Where the licence applies, commercial use is allowed; services above 100 million monthly users need a licence from Tencent. The licence does not apply at all in the European Union, the United Kingdom and South Korea. Read the Tencent Hunyuan Community License itself before using it for work.
Which HunyuanVideo 1.5 file for your card
For each memory size, the highest-precision diffusion file that loads with room to spare (a file that only just fits is used when nothing else does), and what is left for working memory while it samples. "Streams" means even the smallest file is larger than the card: ComfyUI still runs it by streaming weights from system RAM, more slowly. FP4 files are offered only to the Blackwell cards that run them natively.
| Card | Usable | File that loads | Largest stage | Left for the work | Verdict |
|---|---|---|---|---|---|
| RTX 4070 | 12 GB | FP8 scaled | 10.9 GB | 1.1 GB | Loads, tight |
| RTX 5060 Ti | 16 GB | FP8 scaled | 10.9 GB | 5.1 GB | Loads fully |
| RTX 4090 | 24 GB | FP16 | 19.2 GB | 4.8 GB | Loads fully |
| RTX 5090 | 32 GB | FP16 | 19.2 GB | 12.8 GB | Loads fully |
| RTX 6000 Ada | 48 GB | FP16 | 19.2 GB | 28.8 GB | Loads fully |
| H100 SXM | 80 GB | FP16 | 19.2 GB | 60.8 GB | Loads fully |
| RTX PRO 6000 Blackwell | 96 GB | FP16 | 19.2 GB | 76.8 GB | Loads fully |
| DGX Spark | 126 GB | FP16 | 19.2 GB | 106.8 GB | Loads fully |
ComfyUI's template: FP16 diffusion model, FP8 scaled text encoder. Stages: text encoder 9.8 GB, diffusion 19.2 GB. Download 29.0 GB.
Check another card, or pick a file by hand, in the ComfyUI VRAM checker.
Every HunyuanVideo 1.5 file ComfyUI loads
Sizes are the exact byte counts Hugging Face reports. Files marked "template" are the ones ComfyUI's official workflow downloads.
| Part | Precision | Size | File |
|---|---|---|---|
| Diffusion model | FP16 · template | 16.65 GB | hunyuanvideo1.5_720p_t2v_fp16.safetensors |
| Diffusion model | FP8 scaled | 8.33 GB | hunyuanvideo1.5_720p_i2v_cfg_distilled_fp8_scaled.safetensors |
| Text encoder | BF16 | 16.58 GB | qwen_2.5_vl_7b.safetensors |
| Text encoder | FP8 scaled · template | 9.38 GB | qwen_2.5_vl_7b_fp8_scaled.safetensors |
| Glyph encoder | FP16 · template | 0.44 GB | byt5_small_glyphxl_fp16.safetensors |
| VAE | FP16 · template | 2.52 GB | hunyuanvideo15_vae_fp16.safetensors |
What Tencent says it needs
- "Minimum GPU Memory: 14 GB (with model offloading enabled)" (Tencent's own pipeline, with offloading, source)
Those figures and this page measure different things. A vendor's figure covers its own script at its default resolution and length, working memory included. The tables here count the weights each stage holds, which is exact, and leave the working memory as the space that remains, because it depends on your resolution, frame count and attention kernel and nobody publishes it per model. If generation runs out of memory with a file that loads, lower the resolution or the frame count before stepping down a precision. ComfyUI's README says it "can run even the biggest open source models on as low as 4GB vram + 8GB ram" by streaming weights (source), so "streams" means slower, not impossible.
LoRA training for HunyuanVideo 1.5
No trainer we checked publishes a VRAM figure for training HunyuanVideo 1.5 yet (kohya's sd-scripts and musubi-tuner, ostris' ai-toolkit, diffusion-pipe, and Tencent's own repo). When one does, it goes here.
What HunyuanVideo 1.5 is good at, and what to watch
Good at
- Video on mid-range cards: Tencent's own pipeline gives a 14 GB minimum with offloading.
- Text-to-video and image-to-video from the same family.
- Text in the picture: a small glyph encoder rides along with the main text encoder.
Watch out for
- The licence does not apply in the European Union, the United Kingdom or South Korea.
- The 1080p step is a second model the same size as the first, run after it.
- The FP8 file Comfy-Org provides at 720p is the CFG-distilled image-to-video model, not the text-to-video one.
Questions
How much VRAM does HunyuanVideo 1.5 need?
In ComfyUI's own workflow, the largest stage is the diffusion stage at 19.2 GB, and the whole workflow downloads 29.0 GB. Working memory for generation comes on top and grows with resolution and length. Tencent itself says: "Minimum GPU Memory: 14 GB (with model offloading enabled)" (Tencent's own pipeline, with offloading).
Can I run HunyuanVideo 1.5 on 8 GB of VRAM?
Not fully. The smallest file here, FP8 scaled at 8.3 GB, needs 10.9 GB at its largest stage, more than 8 GB holds. ComfyUI will still run it by streaming weights from system RAM, which works but is slower.
Can I run HunyuanVideo 1.5 on 12 GB of VRAM?
Yes, with the FP8 scaled file: its largest stage, the diffusion stage, is 10.9 GB, which loads on 12 GB with 1.1 GB left for working memory, a tight fit. That is little room to generate at full resolution.
Can I run HunyuanVideo 1.5 on 16 GB of VRAM?
Yes, with the FP8 scaled file: its largest stage, the diffusion stage, is 10.9 GB, which loads on 16 GB with 5.1 GB left for working memory.
Which HunyuanVideo 1.5 file should I download?
The highest precision that loads fully on your card. On a 24 GB card that is the FP16 file (16.6 GB). ComfyUI's template downloads FP16 with the FP8 scaled text encoder. FP4 files (NVFP4) are made for Blackwell cards (RTX 50, RTX PRO 6000, DGX Spark); on older cards, use FP8 or INT8.
- Files and sizes: Comfy-Org/HunyuanVideo_1.5_repackaged on Hugging Face, read from the Hugging Face API. Template files: ComfyUI's HunyuanVideo 1.5 guide.
- Model and licence: tencent/HunyuanVideo-1.5 and the Tencent Hunyuan Community License.
- Card memory: manufacturers' specifications, as on each GPU page. The checker covers NVIDIA cards; ComfyUI also runs on AMD and Apple silicon, where file formats and speed differ and we have not sized them.
- Fit rule: a stage loads fully under 85% of usable memory and is tight up to 95%, the same margins the LLM pages use.