Open-weight model
Marigold-v2-Qwen-Image-Edit-2509-NF4
by Fal fal/Marigold-v2-Qwen-Image-Edit-2509-NF4
Marigold-v2-Qwen-Image-Edit-2509-NF4 is an open-weight model from Fal. It has 20.4B parameters. At 16-bit it needs about 49 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
Runs On
What it takes to serve Marigold-v2-Qwen-Image-Edit-2509-NF4 (20.4B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 40.9 GB | 49.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 20.4 GB | 24.5 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 10.2 GB | 12.3 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 23, 2026.
Marigold-v2-Qwen-Image-Edit-2509-NF4 on every accelerator the SAVRN Index prices, at every precision
Model Card
The publisher has not written a card for this model.
Configuration
- Attention heads
- 24
- Quantization
- bitsandbytes
Identity and Version
- Repository
- fal/Marigold-v2-Qwen-Image-Edit-2509-NF4
- Publisher
- Fal
- Task
- Not stated by the source
- Modality
- Other
- Library
- diffusers
- Parameters
- 20.4B parameters
- Languages
- Not stated by the source
- Revision
- c73e0fb6bd864768179937235235ecc31f7d0371
- First published
- 2026-10-05
- Last updated
- 2026-10-05
Files and Weights
6 files, 11.6 GB in total. The weights are 3 files totalling 11.6 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| diffusion_pytorch_model-00001-of-00003.safetensors | Weights | 5.0 GB | 3af374c9a913 |
| diffusion_pytorch_model-00002-of-00003.safetensors | Weights | 5.0 GB | 720d04dee7b0 |
| diffusion_pytorch_model-00003-of-00003.safetensors | Weights | 1.6 GB | 8b6fbad3813c |
| config.json | Configuration | 1.0 KB | — |
| diffusion_pytorch_model.safetensors.index.json | Configuration | 500.9 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 11.6 GB
Released by Fal through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 11.6 GB |
| 16-bit | 40.9 GB |
| 8-bit | 20.4 GB |
| 4-bit | 10.2 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About Marigold-v2-Qwen-Image-Edit-2509-NF4
How much GPU memory does Marigold-v2-Qwen-Image-Edit-2509-NF4 need?
About 49 GB at 16-bit and 12.3 GB at 4-bit: the weights (20.4B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run Marigold-v2-Qwen-Image-Edit-2509-NF4 on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.