Image models · Tencent
HunyuanImage 3.0 (80B-A13B)
- Backbone
- 83.0B
- Native output
- 1024×1024
- Default steps
- 50
- License
- tencent-hunyuan-community
- Released
- 2025-09
How to run it
Original weights · Hugging Face
Full-precision files from Tencent, for fine-tuning, vLLM, Transformers, Diffusers or ComfyUI.
huggingface.co/tencent/HunyuanImage-3.0 (opens in a new tab)
Official repo · GitHub
Reference code, inference scripts and docs from Tencent. The weights themselves download from Hugging Face.
github.com/Tencent-Hunyuan/HunyuanImage-3.0 (opens in a new tab)
git clone https://github.com/Tencent-Hunyuan/HunyuanImage-3.0Memory needed
Peak GPU memory at the native output size. Offloading moves the text encoder, or the whole pipeline, to system RAM between steps.
| Precision | Everything on GPU | Text encoder offloaded | Sequential offload | Download |
|---|---|---|---|---|
FP16 | 170 GB | 170 GB | 61.8 GB | Original weights (opens in a new tab) |
FP8 | 86.9 GB | 86.9 GB | 32.8 GB | Original weights (opens in a new tab) |
Components
- Backbone: 83B
- VAE: 0.1B
Best hardware for HunyuanImage 3.0 (80B-A13B)
The least expensive GPU in our catalog that runs it without offloading is the AMD Radeon PRO W7800 48GB (launch price $2,799) at FP8.
| Quality | Hardware | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA RTX PRO 6000 Blackwell |
FP8 | 86.9 GB / 96.0 GB | 40 s | Runs welltight fit |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA H200 141GB SXM |
FP8 | 86.9 GB / 141 GB | 38 s | Runs well |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA GH200 Grace Hopper (96 GB GPU) |
FP8 | 86.9 GB / 96.0 GB | 38 s | Runs welltight fit |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA B200 |
FP8 | 86.9 GB / 180 GB | 33 s | Runs well |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA GB10 (128 GB) |
FP8 | 86.9 GB / 115 GB | 1.9 min | Slow |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA RTX Spark N1X (6144-core GPU) (128 GB) |
FP8 | 86.9 GB / 102 GB | 2.0 min | Slow |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. |  AMD Ryzen AI Max+ 395 (128 GB) |
FP8 | 86.9 GB / 96.0 GB | 1.7 min | Slowtight fit |
| 90Excellent quality (A+) · 90 of 100: how capable the model is at FP16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. |  AMD Instinct MI325X |
FP16 | 170 GB / 256 GB | 54 s | Slow |
| 90Excellent quality (A+) · 90 of 100: how capable the model is at FP16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip |
FP16 | 170 GB / 252 GB | 61 s | Slow |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. |  NVIDIA H200 NVL |
FP8 | 86.9 GB / 141 GB | 41 s | Slow |
Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.