MOSS-Image GGUF
MOSS-Image-SFT is a text-to-image model from OpenMOSS. It accepts natural-language prompts, tag sequences, and JSON layouts for image composition. This page lists the SFT and 8-step Turbo GGUF files.
MOSS-Image-GGUF on Hugging FaceExample output
Available Quantizations
File sizes use GiB. Standard GGUF quants are common GGML formats. Proprietary ConvRot quants require the comfyui-gguf-reboot loader.
SFT
| Quant | Type | Size | File |
|---|---|---|---|
| Q4_0 | Standard | 3.10 GiB | moss_image_sft-Q4_0.gguf |
| Q4_K_M | Standard | 3.35 GiB | moss_image_sft-Q4_K_M.gguf |
| Q6_K | Standard | 4.24 GiB | moss_image_sft-Q6_K.gguf |
| Q8_0 | Standard | 5.31 GiB | moss_image_sft-Q8_0.gguf |
| Q8_CR | Proprietary, comfyui-gguf-reboot | 5.04 GiB | moss_image_sft_Q8_CR.gguf |
| Q4_CR_L | Proprietary, comfyui-gguf-reboot | 3.56 GiB | moss_image_sft-Q4_CR_L.gguf |
Turbo
The supplied Turbo workflow runs 8 sampling steps.
| Quant | Type | Size | File |
|---|---|---|---|
| Q4_0 | Standard | 3.10 GiB | moss_image_sft_turbo-Q4_0.gguf |
| Q4_K_M | Standard | 3.35 GiB | moss_image_sft_turbo-Q4_K_M.gguf |
| Q6_K | Standard | 4.24 GiB | moss_image_sft_turbo-Q6_K.gguf |
| Q8_0 | Standard | 5.31 GiB | moss_image_sft_turbo-Q8_0.gguf |
| Q8_CR | Proprietary, comfyui-gguf-reboot | 5.04 GiB | moss_image_sft_turbo_Q8_CR.gguf |
| Q4_CR_L | Proprietary, comfyui-gguf-reboot | 3.56 GiB | moss_image_sft_turbo-Q4_CR_L.gguf |
Inputs and outputs
The model accepts natural-language prompts, tag sequences, and JSON layouts that describe objects, text, and their placement. It generates still images. The upstream pipeline requires image dimensions that are multiples of 16. The supplied workflow uses a 1-megapixel portrait canvas by default.
The upstream SFT reference settings use 50 steps and CFG 3. The Turbo workflow uses 8 steps and CFG 1.
| Setting | SFT workflow | Turbo workflow |
|---|---|---|
| Steps | 50 | 8 |
| CFG | 3 | 1 |
| Sampler | euler | euler |
| Scheduler | sgm_uniform | simple |
| Shift | 5 (model default) | 5 (model default) |
| Skip layer guidance | Layer 22, full range | Bypassed, no negative branch at CFG 1 |
ComfyUI setup
Install comfyui-gguf-reboot. Load GGUF files with its UnetLoaderGGUFDynamicVRAM node. These loader nodes give the best performance. Standard GGUF files can be opened with compatible tools, while the custom ConvRot files require this loader.
Download the text encoder and VAE from the Comfy-Org MOSS-Image repository.
| Component | File | ComfyUI folder |
|---|---|---|
| Text encoder | qwen3.5_9b_int8_convrot.safetensors | models/text_encoders/ |
| VAE | flux2-vae.safetensors | models/vae/ |
Workflows
Import these JSON files into ComfyUI.
MOSS-Image SFT, 50 stepsMOSS-Image Turbo, 8 stepsSources and license
The model is based on OpenMOSS-Team/MOSS-Image-SFT. The upstream model is licensed under Apache-2.0. See its model card and license for the original model terms.