Zhenzhen's AI Radar · MOSS Image · 2026-10-29 Evening Report
This issue features 36 new versions: Strongly Recommended 10 · Recommended 17 · Worth Watching 9 · Listed Only 0
Strongly Recommended (10)
[S] TJ Realism Aesthetic · Moss v1.0
LoRA · Ported · Author: vimoi · Published 10-29 16:57 · Downloads 104
To address these, I specifically trained the Moss-version Realism Aesthetic. More realistic lighting and shadow, with improved contrast
Sample Images
Reasons for Recommendation
- Ported old model to MOSS: Previously available in Krea 2 / ZImageTurbo / ZImageBase versions
- Within 4 hours of release, it has already received 104 downloads and 10 likes, with a download rate higher than 99%% of MOSS models from the last 7 days
- Includes 12 sample images, allowing you to evaluate the results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows image generation, on-site rental, and rental
How to Use
- Load this LoRA in the MOSS workflow (277 MB · BF16) and include the trigger word "virealism" in the prompt
- Author's parameter: "Weight: 0.8-1"
- Author's parameter: "Steps: 25"
Risks and Limitations
- 8 sensitive-level examples are hidden
Trigger Word: virealism
[S] Chibi Head | ChibiHead | Moss image · v1.0
LoRA · New Model · Author: Akai_azusa · Published 10-29 17:15 · Downloads 37
Minimalist chibi head style LoRA. The native Moss base model can also generate this chibi head style, but it requires a very long prompt.
Sample Images
Reasons for Recommendation
- Within 4 hours of release, it has already received 37 downloads and 5 likes, with a download rate higher than 95%% of MOSS models from the last 7 days
- Includes 11 sample images, allowing you to evaluate the results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows on-site rental and rental
How to Use
- Load this LoRA in the MOSS workflow (95 MB · BF16) and include the trigger word "chibi_head" in the prompt
- Author's parameter: "Parameters: MOSS-Image Base, CFG 3, Shift 5, 1024×1024, 20 to 50 steps, euler + simple."
- Author's parameter: "| Steps | 20 – 50 |"
Trigger Word: chibi_head
[S] FasciumMOSS SFT ARTY · v2.0
Base Model / Checkpoint · Version Update · Author: FASCIUM · Published 10-29 13:54 · Downloads 44
Unlike typical LoRA merges that layer styles on top of a base, this model has been deeply trained to understand art history at a structural level.
Sample Images
Reasons for Recommendation
- The 4th MOSS version of "FasciumMOSS SFT ARTY"
- Within 7 hours of release, it has already received 44 downloads and 10 likes, with a download rate higher than 91%% of MOSS models from the last 7 days
- Includes 10 sample images, allowing you to evaluate the results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows image generation, on-site rental, rental, sale, and sale of merges
How to Use
- Full weights ?, replace MOSS base model for loading
- Author's parameter: "Set sampler to Euler a , scheduler to Tangent , steps to 50 , CFG to 3.0"
- Author's parameter: "Experiment with resolution : Portrait (768×1152) works best for figures, square (1024×102…"
Risks and Limitations
- 9 sensitive-level examples are hidden
- Version notes (update content) are empty
[S] Seapall - Artist Style · MOSS-Image
LoRA · Ported · Author: Azriaelis · Published 10-29 09:30 · Downloads 66
Soft painterly anime style with refined linework, atmospheric glow, delicate texture, semi-realistic shading, elegant detailing, and cinematic contrast.
Sample Images
Reasons for Recommendation
- Ported old model to MOSS: Previously available in Anima / Krea 2 versions
- Within 12 hours of release, it has already received 66 downloads and 9 likes, with a download rate higher than 90%% of MOSS models from the last 7 days
- Includes 10 sample images, allowing you to evaluate the results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows image generation and on-site rental
How to Use
- Load this LoRA in the MOSS workflow (246 MB · BF16) and include the trigger words "@seapall", "gritty", and "sketchy" in the prompt
- Author's parameter: "Lora Strength: 0.6-0.8 | Steps: 25 | CFG: 4 Recommended settings for Krea2Turbo:"
- Author's parameter: "Lora Strength: 0.8-1.4 | Steps: 8-10 | CFG: 1 Recommended settings for MOSS-Image:"
Risks and Limitations
- 8 sensitive-level examples are hidden
Trigger Words: @seapall, gritty, sketchy
[S] waxa Cypressionism · v2.0 Moss Base
LoRA · New Model · Author: superkaero · Published 10-29 10:56 · Downloads 53
waxa Cypressionism is a style LoRA trained based on Moss Image Base. Its true value lies in "stacking": when combined with other anime or painting LoRAs, it acts as a layer of color and brushstroke tinting, often creating effects that neither could achieve alone.
Sample Images
Reasons for Recommendation
- Within 10 hours of release, it has already received 53 downloads and 9 likes, with a download rate higher than 88%% of MOSS models from the last 7 days
- Includes 13 sample images, allowing you to evaluate the results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows image generation, on-site rental, rental, sale, and sale of merges
How to Use
- Load this LoRA in the MOSS workflow (71 MB · BF16); the author has not registered a trigger word, refer to example prompts for effect description
- Author's parameter: "Recommended strength: 1.0–2.0"
- Author's parameter: "Combine with 8-step acceleration LoRA for quick sample generation"
Risks and Limitations
- 7 sensitive-level examples are hidden
- No trigger word registered
[S] Azriael's Niji Shadow · MOSS-Image
LoRA · Ported · Author: Azriaelis · Published 10-29 11:00 · Downloads 27
Semi-realistic Niji style trained on my Niji style.
Sample Images
Reasons for Recommendation
- Ported old model to MOSS: Previously available in Krea 2 version
- Within 10 hours of release: 27 downloads, 4 likes
- Includes 12 sample images, allowing you to evaluate the results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows image generation and on-site rental
How to Use
- Load this LoRA in the MOSS workflow (246 MB · BF16) and include the trigger word "nijishadow" in the prompt
- Author's parameter: "Lora Strength: 0.8-1.4 | Steps: 8-10 | CFG: 1 Recommended settings for MOSS-Image:"
- Author's parameter: "Base with 8step turbo lora"
Risks and Limitations
- 8 sensitive-level examples are hidden
Trigger Word: nijishadow
[S] Realism By Stable Yogi Moss · v1.5 INT6
Base Model / Checkpoint · Version Update · Author: Stable_Yogi · Published 10-29 20:30 · Downloads 29
Realism by Stable Yogi · Moss · V1.5 Real-looking photos on the Moss engine — now steadier on busy scenes and hard poses.
Sample Images
Reasons for Recommendation
- The 6th MOSS version of "Realism By Stable Yogi Moss", this update: INT6-W6A8 — 4.5 GB — near-INT8 quality in a smaller file. Real-looking photos on the Moss engine — now steadier on busy scenes and hard poses.
- Within 1 hours of release, it has already received 29 downloads and 4 likes, with a download rate higher than 97%% of MOSS models from the last 7 days
- Includes 2 sample images, allowing you to evaluate the results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows image generation and on-site rental
How to Use
- Full weights 4.2 GB · SafeTensor, replaces MOSS base model loading
- Author parameters: “Same speed: full images in 8 steps .”
- Author parameters: “10 steps · CFG 1 · ER SDE · Beta · 1024×1536.”
Risks and Limitations
- 18 sensitive-level examples hidden
[S] [MOSS Image] d3style anime illustration style · v1
LoRA · New Model · Author lepetitprince716355 · Published 10-29 20:18 · Downloads 15
An anime illustration style LoRA for MOSS-Image: larger eyes with heavy lashes, a soft blush under the eyes, glossy hair and warm, deep shading.
Images
Reasons for Recommendation
- Published 1 hours ago with 15 downloads and 0 likes; download speed is higher than 92% of MOSS models from the last 7 days
- Includes 6 example images, allowing you to preview results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier for reproduction
- Free download; license permits image generation, on-site rental, and rental
How to Use
- Load this LoRA in the MOSS workflow (142 MB · BF16) and include the trigger word “d3style” in the prompt
- Author parameters: “First release: the step 2000 checkpoint.”
- Author parameters: “Strength 0.6 keeps the base model's outfit and figure;”
Trigger Words: d3style
[S] Genshin Vodyanitsa | Genshin Impact Vodyanitsa · v1.0
LoRA · New Model · Author Suzukanaoko · Published 10-29 16:47 · Downloads 26
A character LoRA for Vodyanitsa from Genshin Impact, compatible with both Base and SFT base models. If you only write the name, the bangs may sometimes appear thin; to consistently generate thick bangs, include “Vodyanitsa thick bangs” in the prompt.
Images
Reasons for Recommendation
- Published 4 hours ago with 26 downloads and 4 likes; download speed is higher than 90% of MOSS models from the last 7 days
- Includes 2 example images, allowing you to preview results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier for reproduction
- Free download; license permits image generation, on-site rental, rental, sales, and sale of fused models
How to Use
- Load this LoRA in the MOSS workflow (189 MB · BF16) and include the trigger words “vodyanitsa (genshin impact)” and “Vodyanitsa” in the prompt
- Author parameters: “Standard generation: 30–50 steps, CFG 3–3.5, euler / simple, shift 5, LoRA weight 1.0”
- Author parameters: “DMD acceleration: 8–12 steps, CFG 1, euler / simple, shift 5, LoRA weight 1.0”
Risks and Limitations
- 6 sensitive-level examples hidden
Trigger Words: vodyanitsa (genshin impact), 沃雅妮莎
[S] Vladimir Zarubin — The Master of the Soviet Greeting Postcard | The Whole Holiday Calendar in Gouache | LoRA (Krea2 Turbo) · MOSS
LoRA · Ported · Author FASCIUM · Published 10-29 13:12 · Downloads 31
Vladimir Zarubin — The Master of the Soviet Greeting Postcard | The Whole Holiday Calendar in Gouache | LoRA (Krea2 Turbo)
Images
Reasons for Recommendation
- Ported old model to MOSS: a Krea 2 version previously existed
- Published 8 hours ago with 31 downloads and 11 likes; download speed is higher than 83% of MOSS models from the last 7 days
- Includes 3 example images, allowing you to preview results before deciding to download
- The author provides specific parameters (2 instances), lowering the barrier for reproduction
- Free download; license permits image generation, on-site rental, rental, sales, and sale of fused models
How to Use
- Load this LoRA in the MOSS workflow (123 MB · BF16); the author has not registered trigger words, so refer to example prompts to describe effects
- Author parameters: “Not required at CFG 1.”
- Author parameters: “stable at weight 0.9–1.0.”
Risks and Limitations
- 1 sensitive-level examples hidden
- No registered trigger words
Recommendations (17)
[A] MOSS-FET · v0.1
Base Model / Checkpoint · New Model · Author aueki4g467 · Published 10-29 10:54 · Downloads 93
Since support for MOSS-Image in OneTrainer—the tool I usually use—is lagging, I’m testing a full fine-tuning (FT) process using a different tool.
Images
Reasons for Recommendation
- Published 10 hours ago with 93 downloads and 13 likes; download speed is higher than 94% of MOSS models from the last 7 days
- Includes 12 example images, allowing you to preview results before deciding to download
- Description is relatively complete (approx. 1161 words)
- Free download; license permits image generation, on-site rental, and sale of fused models
How to Use
- Full weights 5.0 GB · INT8 · SafeTensor, replaces MOSS base model loading
Risks and Limitations
- 8 sensitive-level examples hidden
[A] University Classroom — Japanese / American / Chinese / Lecture Hall, CCTV & Eye-Level | MOSS · v1.0
Base Model / Checkpoint · New Model · Author 858654128784 · Published 10-29 17:30 · Downloads 65
Japanese-style: single wooden desks, a broad green chalkboard, a full row of windows with cream curtains on the left, glazed sliding corridor doors on the righ…
Images
Reasons for Recommendation
- Published 4 hours ago with 65 downloads and 13 likes; download speed is higher than 98% of MOSS models from the last 7 days
- Includes 4 example images, allowing you to preview results before deciding to download
- The author provides specific parameters (1 instances), lowering the barrier for reproduction
- Free download; license permits image generation, on-site rental, rental, sales, and sale of fused models
How to Use
- Full weights 142 MB · FP16 · SafeTensor, replaces MOSS base model loading
- Author parameters: “Recommended weight 0.6.”
Risks and Limitations
- 16 sensitive-level examples hidden
[A] Shoes & Socks Realism — Sneakers, Sports Socks, Worn & Stuffed | MOSS · v1.0
LoRA · New Model · Author 858654128784 · Published 10-29 12:05 · Downloads 55
Basketball socks on feet: many colors and stripes, seated or standing, feet on wood or tile.
Images
Reasons for Recommendation
- Published 9 hours ago with 55 downloads and 6 likes; download speed is higher than 91% of MOSS models from the last 7 days
- Includes 12 example images, allowing you to preview results before deciding to download
- The author provides specific parameters (1 instances), lowering the barrier for reproduction
- Free download; license permits image generation, on-site rental, rental, sales, and sale of fused models
How to Use
- Load this LoRA in the MOSS workflow (142 MB · FP16); the author has not registered trigger words, so refer to example prompts to describe effects
- Author parameters: “Recommended weight 0.6.”
Risks and Limitations
- 8 sensitive-level examples hidden
- No registered trigger words
[A] Moody Moss Mix (Free download 免费下载) · V3.3 25STEPS CFG3
Base Model / Checkpoint · Version Update · Author catlover1937 · Published 10-29 09:05 · Downloads 691
🎨 I will continue to share more high-quality content for free ✨, please support me by buying me a coffee! 🤔 The Story So Far: Why Did V1 & V2 Turn Everyone Asian?
Images
Reasons for Recommendation
- The 7th MOSS version of “Moody Moss Mix (Free download 免费下载)”
- Published 12 hours ago with 691 downloads and 58 likes; download speed is higher than 100% of MOSS models from the last 7 days
- Includes 1 example images, allowing you to preview results before deciding to download
- The author provides specific parameters (3 instances), lowering the barrier for reproduction
- Free download; license permits on-site rental
How to Use
- Full weights 4.7 GB · FP8 · SafeTensor, replaces MOSS base model loading
- Author parameters: “The custom LoRA blend is now finely tuned specifically for 25 steps.”
- Author parameters: “Steps: 25 (Optimized for this speed)”
Risks and Limitations
- 19 sensitive-level examples hidden
- Version notes (update content) are empty
[A] ciTron Legacy (World Morph) [MOSS & ZIT & Pony] · MOSS
LoRA · Port · Author CitronLegacy · Published 10-29 20:44 · Downloads 6
I'm proud to formally introduce myself! My name is Citron Legacy and this is my home world.
Image
Reasons to Recommend
- Ported old model to MOSS: Previous versions existed for ZImageTurbo / Pony / SD 1.5
- Published 1 hours ago: 6 downloads, 2 likes
- Includes 15 example images, allowing you to check the results before downloading
- Documentation is fairly complete (approx. 440 words)
- Free download, license allows image generation, in-site rental, and rental
How to Use
- Load this LoRA in the MOSS workflow (142 MB · BF16), and include the trigger word "c1tr0nL" in the prompt; the author also provides complete example prompts, see the copyable tag below
Risks and Limitations
- 5 sensitive-level examples have been hidden
Trigger Words: c1tr0nL,, c1tr0nL, science fiction, glowing, neon trim, neon,
【A】V67 Ghibli Look · V1 MOOS
LoRA · Port · Author angelomaiota · Published 10-29 13:56 · Downloads 31
It does not turn your photos into cartoons. This LoRA teaches FLUX to see the world through Miyazaki's eyes.
Image
Reasons to Recommend
- Ported old model to MOSS: Previous versions existed for Flux.1 D / Krea 2 / ZImageTurbo
- Published 7 hours ago, already has 31 downloads and 3 likes; download speed is higher than 84% of MOSS models from the last 7 days
- Includes 3 example images, allowing you to check the results before downloading
- The author provides specific parameters (2 instances), lowering the barrier to reproduction
- Free download, license allows image generation, in-site rental, rental, sale, and sale of fusions
How to Use
- Load this LoRA in the MOSS workflow (123 MB · BF16); the author has not registered a trigger word, refer to the example prompts to describe the effect
- Author's parameters: "LoRA Weight: 0.6 - 0.8 (balances photorealism with Ghibli essence)"
- Author's parameters: "CFG Scale: 3.5 - 5.0 (lower for softer, dreamier results)"
Risks and Limitations
- 1 sensitive-level examples have been hidden
- No trigger word registered
【A】Genshin Impact Miliastra Wonderland - Fireflies of Hope | Starry Realm - Good Time Fireflies · v1.0 MOSS
LoRA · Port · Author LS3171244 · Published 10-29 11:30 · Downloads 22
Recommended strength:1.0
Image
Reasons to Recommend
- Ported old model to MOSS: Previous versions existed for Anima / Illustrious
- Published 10 hours ago: 22 downloads, 1 likes
- Includes 11 example images, allowing you to check the results before downloading
- The author provides specific parameters (2 instances), lowering the barrier to reproduction
- Free download, license allows in-site rental and rental
How to Use
- Load this LoRA in the MOSS workflow (142 MB · BF16), and include the trigger words "F1r3fl13s0fH0p3", "beret", and "blue footwear" in the prompt; the author also provides complete example prompts, see the copyable tag below
- Author's parameters: "Recommended weight 1.0"
- Author's parameters: "Recommended strength:1.0"
Risks and Limitations
- 1 sensitive-level examples have been hidden
- Description is too short, usage is mainly judged by examples
Trigger Words: F1r3fl13s0fH0p3, long sleeves, capelet, short over long sleeves, dress, white pantyhose, beret, white headwear, blue footwear
【A】4BIT ALCHEMY-16-Colour Pixel Art · v1.0
LoRA · New Model · Author Addictedd · Published 10-29 18:19 · Downloads 34
A 4-bit pixel-art style Chunky blocks, hard edges, banded skies and a fixed 16-colour palette.
Image
Reasons to Recommend
- Published 3 hours ago, already has 34 downloads and 7 likes; download speed is higher than 97% of MOSS models from the last 7 days
- Includes 5 example images, allowing you to check the results before downloading
- Free download, license allows image generation, in-site rental, rental, sale, and sale of fusions
How to Use
- Load this LoRA in the MOSS workflow (66 MB · BF16), and include the trigger word "pxart" in the prompt
Risks and Limitations
- 1 sensitive-level examples have been hidden
Trigger Words: pxart
【A】@02charaice20 2026/2025 Style · moss_v0.0.0
LoRA · New Model · Author wraith_executioner_year_2 · Published 10-29 11:56 · Downloads 33
Captioning: Booru tags only.
Image
Reasons to Recommend
- Published 9 hours ago, already has 33 downloads and 7 likes; download speed is higher than 79% of MOSS models from the last 7 days
- Includes 4 example images, allowing you to check the results before downloading
- The author provides specific parameters (1 instances), lowering the barrier to reproduction
- Free download, license allows image generation, in-site rental, rental, sale, and sale of fusions
How to Use
- Load this LoRA in the MOSS workflow (123 MB · BF16), and include the trigger words "@02charaice20", "year 2026", and "year 2025" in the prompt; the author also provides complete example prompts, see the copyable tag below
- Author's parameters: "LORA, MOSS Image Base, dim32/alpha32, AdamW8bit, LR 1e-4, cosine 1900steps, 108images, 10…"
Risks and Limitations
- 2 sensitive-level examples have been hidden
Trigger Words: @02charaice20, pixel art, year 2026, year 2025
【A】xSkyler · MOSS Image
LoRA · New Model · Author randomimagepotato · Published 10-29 19:23 · Downloads 29
Once my buzz builds back up I will make a Krea2 version. 100% AI generated, no real pictures were used.
Image
Reasons to Recommend
- Published 2 hours ago, already has 29 downloads and 3 likes; download speed is higher than 97% of MOSS models from the last 7 days
- Includes 3 example images, allowing you to check the results before downloading
- Documentation is fairly complete (approx. 752 words)
- Free download, license allows image generation and in-site rental
How to Use
- Load this LoRA in the MOSS workflow (123 MB · BF16), and include the trigger words "xSkyler" and "a 21 year old adult woman with long natural glossy blonde curly hair" in the prompt
Risks and Limitations
- 11 sensitive-level examples have been hidden
Trigger Words: xSkyler, a 21 year old adult woman with long natural glossy blonde curly hair
【A】redi (rasec asdjh) - Artist style MOSS · Redi (rasec asdjh)
LoRA · New Model · Author empek17 · Published 10-29 12:18 · Downloads 45
Style of redi (rasec asdjh). It's quite literally my first lora but i think it turned out decent.
Image
Reasons to Recommend
- Published 9 hours ago, already has 45 downloads and 6 likes; download speed is higher than 87% of MOSS models from the last 7 days
- Includes 2 example images, allowing you to check the results before downloading
- The author provides specific parameters (3 instances), lowering the barrier to reproduction
- Free download, license allows image generation, in-site rental, and rental
How to Use
- Load this LoRA in the MOSS workflow (142 MB · BF16); the author has not registered a trigger word, refer to the example prompts to describe the effect
- Author's parameters: "Base with 8step turbo lora"
- Author's parameters: "Lora Strength: 1"
Risks and Limitations
- 3 sensitive-level examples have been hidden
- No trigger word registered
【A】Player Profile — Football Player Page With Absurd Stats | MOSS · v1.0
LoRA · New Model · Author 858654128784 · Published 10-29 15:23 · Downloads 52
Translucent panels and thin dividers, like a real sports app. Chinese and English UI both work.
Image
Reasons to Recommend
- Published 6 hours ago, already has 52 downloads and 8 likes; download speed is higher than 95% of MOSS models from the last 7 days
- Includes 3 example images, allowing you to check the results before downloading
- Documentation is fairly complete (approx. 884 words)
- Free download, license allows image generation, in-site rental, rental, sale, and sale of fusions
How to Use
- Load this LoRA in the MOSS workflow (142 MB · FP16); the author has not registered a trigger word, refer to the example prompts to describe the effect
Risks and Limitations
- 17 sensitive-level examples hidden
- No registered trigger words
【A】Tales of Old Rus · MOSS Image.V1.0
LoRA · Port · Author GreenRazor · Published 10-29 12:33 · Downloads 31
An impressive LoRA model inspired by the world of "Tales of Ancient Rus" and Slavic fantasy, designed for an adult audience.
Images
Reasons to Recommend
- Ported old model to MOSS: An SDXL 1.0 version existed previously
- 9 hours after release, it has 31 downloads and 5 likes, with a download rate higher than 80%% of MOSS models from the last 7 days
- Includes 3 example images, allowing you to check the results before deciding to download
- The author provides specific parameters (1 instances), lowering the barrier to reproduction
- Free download, license allows image generation, on-site rental, rental, sale, and sale of merges
How to Use
- Load this LoRA in the MOSS workflow (142 MB · BF16); the author has not registered trigger words, refer to the example prompts to describe the effect
- Author's parameters: “I recommend using a weight of 0.5.”
Risks and Limitations
- 2 sensitive-level examples hidden
- No registered trigger words
【A】Fire Station Garage — Fire Engine Bay, CCTV & Eye-Level | MOSS · v1.0
LoRA · New Model · Author 858654128784 · Published 10-29 10:55 · Downloads 55
Red fire engines in yellow-lined bays with chrome grilles and bumpers, roof light bars and folded hoses on top, open side equipment compartments and silver dia…
Images
Reasons to Recommend
- 10 hours after release, it has 55 downloads and 11 likes, with a download rate higher than 89%% of MOSS models from the last 7 days
- Includes 4 example images, allowing you to check the results before deciding to download
- Description is relatively complete (approx. 1587 characters)
- Free download, license allows image generation, on-site rental, rental, sale, and sale of merges
How to Use
- Load this LoRA in the MOSS workflow (142 MB · FP16); the author has not registered trigger words, refer to the example prompts to describe the effect
Risks and Limitations
- 15 sensitive-level examples hidden
- No registered trigger words
【A】Cora NordhavenLove · MossImage
LoRA · Port · Author MisterMR · Published 10-29 14:00 · Downloads 28
This LoRA does not depict a real person;
Images
Reasons to Recommend
- Ported old model to MOSS: Krea 2 / MiniMax H3 versions existed previously
- 7 hours after release, it has 28 downloads and 2 likes, with a download rate higher than 83%% of MOSS models from the last 7 days
- Includes 5 example images, allowing you to check the results before deciding to download
- Description is relatively complete (approx. 2990 characters)
- Free download, license allows on-site rental, rental, and image generation
How to Use
- Load this LoRA in the MOSS workflow (123 MB); the author has not registered trigger words, refer to the example prompts to describe the effect
Risks and Limitations
- 15 sensitive-level examples hidden
- No registered trigger words
【A】ComfyUI-MOSS-QwenVL-Layout · v1.1.0
Workflow · Version Update · Author leroy989 · Published 10-29 10:17 · Downloads 85
A ComfyUI node that turns a reference image into a MOSS-Image layout prompt . Qwen3-VL / Qwen3.5-VL looks at your image, finds the things you ask for (includin…
Images
Reasons to Recommend
- The 2th MOSS version of "ComfyUI-MOSS-QwenVL-Layout", this update: New prompt only mode that will take a basic prompt and turn that into a full bbox guided prompt. Built image scaling into the node.
- 11 hours after release, it has 85 downloads and 10 likes, with a download rate higher than 93%% of MOSS models from the last 7 days
- Includes 2 example images, allowing you to check the results before deciding to download
- Description is relatively complete (approx. 936 characters)
- Free download, license allows image generation, on-site rental, rental, sale, and sale of merges
How to Use
- ComfyUI workflow: comfyuiMOSSQwenvl_v110.json (51 KB); after importing, follow the instructions to add custom nodes and models, it is not a new weight itself
Risks and Limitations
- 2 sensitive-level examples hidden
- Dependent nodes and models need to be verified by yourself, this tool has not been run
【A】Kzmynon AI Style [K/A] · moss-v2
LoRA · Version Update · Author Peslanna · Published 10-29 12:39 · Downloads 66
Highly saturated colors and skin texture Moss v1 performs better physically, while moss v2 is relatively a bit darker in terms of visuals.
Images
Reasons to Recommend
- The 2th MOSS version of "Kzmynon AI Style [K/A]"
- 8 hours after release, it has 66 downloads and 7 likes, with a download rate higher than 93%% of MOSS models from the last 7 days
- Includes 2 example images, allowing you to check the results before deciding to download
- The author provides specific parameters (2 instances), lowering the barrier to reproduction
- Free download, license allows on-site rental and rental
How to Use
- Load this LoRA in the MOSS workflow (142 MB · BF16); the author has not registered trigger words, refer to the example prompts to describe the effect
- Author's parameters: “strength 1.2 or lower”
- Author's parameters: “Anyway, just use turbo Lora , or Base\sft 50 steps, 3 cfg”
Risks and Limitations
- 1 sensitive-level examples hidden
- No registered trigger words
- Version notes (update content) are empty
Worth Following (9)
【B】Storyflow Character Card · MOSS Workflow Beta - Four Views - Privacy Toggle
LoRA · Version Update · Author Coffmeteor · Published 10-29 16:27 · Downloads 51
This update adds a fourth panel for an adult front-facing body-proportion reference. Added a native privacy_blocks On/Off control, enabled by default.
Images
【B】KreLuminA · moss-v1.25-BPP
LoRA · Version Update · Author henxpoi521 · Published 10-29 18:09 · Downloads 25
Merged with Krea2 to generate semi-realistic images with highlights in the prompts, but the hair is completely messy, use with caution. I tried AI-Toolkit's Blank Prompt Preservation (BPP)
Images
【B】Moody Moss Mix (Free download 免费下载) · V3.3 25STEPS CFG3 BF16
Base Model / Checkpoint · Version Update · Author catlover1937 · Published 10-29 09:14 · Downloads 2 · Paid
🎨 I will continue to share more high-quality content for free✨, welcome to support me, buy me a coffee! 🤔 The Story So Far: Why Did V1 & V2 Turn Everyone Asian?
Images
【B】ModelingClay [Minimax / MOSS / Krea2] · MOSS v1.0
LoRA · Port · Author TeraWatt · Published 10-29 09:57 · Downloads 3 · Paid
This is a style LoRA model designed to replicate characters or figures made of modeling clay, as if they were intended for a stop-motion production.
Images
【B】qaq moss image · v1.0
LoRA · New Model · Author Eustoma2 · Published 10-29 16:14 · Downloads 52
Old art style, trigger word QAQSTYLE, effectively improves texture, recommended 0.6
Images
【B】ZAKHAR — Dark-Realism Chiaroscuro | LoRA (Krea2 Turbo) · MOSS
LoRA · Port · Author FASCIUM · Published 10-29 12:56 · Downloads 46
ZAKHAR — Dark-Realism Chiaroscuro | LoRA (Krea2 Turbo) Mood: introspection through minimalism and controlled air .
Images
【B】deepseek whale girl · v1.0
LoRA · New Model · Author koivis · Published 10-29 18:22 · Downloads 15
@deepseek, an anime-style illustration of a blue whale girl with long wavy blue hair, pointed whale-fin ears with blue bows, and a big dark navy whale tail, <服…
Image
【B】Oil Painting [ Minimax / MOSS / Krea2]🎨 🖌️ 🖼️ · MOSS v1.0
LoRA · Ported · Author TeraWatt · Published 10-29 12:41 · Downloads 0 · Paid
This is a finely tuned model, trained on a carefully curated dataset, designed to replicate brushstrokes and the physical volume of oil paint—complete with ric…
Image
【B】Pose Loader · v1.0
Workflow · New Model · Author Qingyunluo924 · Published 10-29 12:26 · Downloads 43
Created a ComfyUI node based on the original author's content. It is used solely for loading images. You can use LLM models to reverse-engineer prompts or perform image-to-image generation. Give it a try. Download and extract it to the comfyui custom_nodes directory to use.
Image
Data: Civitai Public API (Base Model MOSS Image), collection window 10-29 08:31 → 10-29 21:01 (Beijing Time). Recommendation reasons are automatic drafts; download counts only reflect attention levels.