工作流 /
图像 /
Ming Image 0.1 设计：文生图
Ming Image 0.1 设计：文生图
Ming Image 0.1 Design: Text to Image
使用 Ming Image 0.1 Design 从文本提示生成文本密集型设计作品，这是一个针对 UI 屏幕、信息图表和海报调优的 6B 模型。它不接受文件输入，仅生成一张图像，支持透明背景以便于分层。非常适合界面原型、营销海报和数据丰富的信息图表草稿。
图像
5 个节点 20 次使用 ComfyUI ≥ 0.44.0 ComfyUI 69 KB
需要的权重
这个工作流引用了 4 个权重文件，其中 4 个能在本站模型库里找到。
ming_image_0.1_design_int8_convrot.safetensors
在模型库查看
ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors
在模型库查看
ming_image_vae_bf16.safetensors
在模型库查看
qwen3.8_27b_w4a8.safetensors
在模型库查看
节点构成
MarkdownNote
× 2
ResolutionSelector
× 1
4d166aa4-f08c-43b7-b889-2b43e3b4c535
× 1
SaveImageAdvanced
× 1
官方标注的模型
Ming Image
工作流 JSON
和上面「下载工作流 JSON」是同一份文件。
1{
2 "id": "29844965-cf09-4b2e-ad35-4e07b362a675",
3 "revision": 0,
4 "last_node_id": 85,
5 "last_link_id": 180,
6 "nodes": [
7 {
8 "id": 75,
9 "type": "ResolutionSelector",
10 "pos": [
11 -1030,
12 4130
13 ],
14 "size": [
15 270,
16 150
17 ],
18 "flags": [],
19 "order": 0,
20 "mode": 0,
21 "showAdvanced": false,
22 "inputs": [],
23 "outputs": [
24 {
25 "name": "width",
26 "type": "INT",
27 "links": [
28 163
29 ]
30 },
31 {
32 "name": "height",
33 "type": "INT",
34 "links": [
35 164
36 ]
37 }
38 ],
39 "properties": {
40 "Node name for S&R": "ResolutionSelector"
41 },
42 "widgets_values": [
43 "1:1 (Square)",
44 1,
45 8
46 ],
47 "widgets_values_named": {
48 "aspect_ratio": "1:1 (Square)",
49 "megapixels": 1,
50 "multiple": 8
51 }
52 },
53 {
54 "id": 80,
55 "type": "4d166aa4-f08c-43b7-b889-2b43e3b4c535",
56 "pos": [
57 -710,
58 3860
59 ],
60 "size": [
61 470,
62 570
63 ],
64 "flags": [],
65 "order": 3,
66 "mode": 0,
67 "inputs": [
68 {
69 "label": "prompt",
70 "name": "value_2",
71 "type": "STRING",
72 "widget": {
73 "name": "value_2"
74 },
75 "link": null
76 },
77 {
78 "label": "width",
79 "name": "value",
80 "type": "INT",
81 "widget": {
82 "name": "value"
83 },
84 "link": 163
85 },
86 {
87 "label": "height",
88 "name": "value_1",
89 "type": "INT",
90 "widget": {
91 "name": "value_1"
92 },
93 "link": 164
94 },
95 {
96 "label": "enable_PE",
97 "name": "switch",
98 "type": "BOOLEAN",
99 "widget": {
100 "name": "switch"
101 },
102 "link": null
103 },
104 {
105 "label": "PE_model",
106 "name": "clip_name_1",
107 "type": "COMBO",
108 "widget": {
109 "name": "clip_name_1"
110 },
111 "link": null
112 },
113 {
114 "label": "attention_backend",
115 "name": "attention",
116 "type": "COMBO",
117 "widget": {
118 "name": "attention"
119 },
120 "link": null
121 }
122 ],
123 "outputs": [
124 {
125 "name": "IMAGE",
126 "type": "IMAGE",
127 "links": [
128 176
129 ]
130 }
131 ],
132 "properties": {
133 "previewExposures": []
134 },
135 "widgets_values": [
136 "A retro 1950s mid-century e-commerce homepage design, full-page web layout, square canvas, vintage printed-paper aesthetic: aged cream background with fine halftone grain, slightly yellowed edges, thin dark-green hairline rules separating the sections, muted palette of deep pine green, teal, coral red-orange, warm red and off-white.\n\nTop: a full-width muted teal announcement bar with centered small uppercase serif text \"★ FREE SHIPPING ON ORDERS OVER $75 ★\", with a tiny star at each end. Below it the header, laid out on one row: on the left a horizontal menu in small dark-green uppercase condensed type reading \"NEW ARRIVALS WOMEN MEN ACCESSORIES SALE ABOUT\"; centered, the brand block with the tiny line \"EST. ★ 2010\" above a large warm-red hand-painted script wordmark \"Timeless\" with a thin off-white outline, then \"VINTAGE\" in bold dark-green serif capitals beneath it and a small \"CLOTHING CO.\" line with flanking rules; on the right three small dark-green line icons with tiny uppercase labels \"SEARCH\", \"ACCOUNT\", \"CART\", the cart icon in coral red. A horizontal hairline rule runs under the header.\n\nHero: a large rounded rectangle panel in pale mint with a thin dark-green border, containing on the left the small uppercase line \"★ AUTHENTIC STYLES. TIMELESS YOU.\", then the word \"Vintage\" in enormous warm-red hand-painted script with a soft off-white outline and a subtle extruded shadow, then \"NEVER FADES\" in heavy dark-green serif capitals, then two lines of centered grey-green body copy \"Handpicked vintage clothing from every era. Unique pieces with a story to tell.\", then a coral ribbon-shaped button with a notched tail and small star at each end containing the uppercase text \"SHOP NEW ARRIVALS →\". On the right side of the panel: a 1950s woman in a pale yellow cardigan, plaid full skirt, sunglasses and a polka-dot headscarf standing beside a pink 1950s sedan, plus a leaning vintage roadside sign painted in red and dark green with the text \"Timeless\" in script and \"VINTAGE CLOTHING CO.\" in capitals, and a small teal starburst badge to the right reading \"SUSTAINABLE FASHION TIMELESS STYLE\" in tiny stacked uppercase type.\n\nMiddle: a centered section heading \"★ SHOP BY CATEGORY ★\" in bold dark-green uppercase with small coral stars. Beneath it a single row of five circular product medallions, each a vintage photograph inside a cream ring with a thin dark-green outline: a smiling woman in a pink dress, a man in a brown suit, a brown leather handbag, a pair of dark loafers, and a blue denim jacket. Under each medallion sits a coral ribbon banner with a notched tail holding the uppercase label \"★ DRESSES ★\", \"★ MEN ★\", \"★ ACCESSORIES ★\", \"★ SHOES ★\", \"★ OUTERWEAR ★\", each with a small green star below the ribbon.\n\nLower band: a wide pale mint panel with a thin border. Inside on the left, a woman in a green blouse playing a vinyl record on a cream and brown suitcase-style record player. On the right, a warm photograph of a small vintage storefront with a striped awning and signs reading \"Timeless\" and \"VINTAGE FINDS\". Centered between them: \"More Than Clothes,\" in dark-green hand-written script, then \"★ IT'S A LIFESTYLE ★\" in bold dark-green uppercase, then three lines of grey-green body copy \"We believe in quality, individuality, and the charm of the past. Every piece is curated with care and meant to be loved for years to come.\", then a coral ribbon button reading \"READ OUR STORY →\".\n\nFooter: a darker teal band across the bottom with a small envelope icon and the heading \"JOIN OUR MAILING LIST\" beside the line \"Get updates on new arrivals and special offers!\", three small circular social icons in the middle (facebook, instagram, pinterest) under a thin vertical hairline, a heading \"CUSTOMER CARE\" with the items \"Shipping\", \"Returns\", \"FAQ\", \"Contact Us\", and on the right \"© 2024 TIMELESS VINTAGE\" with \"All rights reserved.\" beneath it.\n\nStyle notes: authentic mid-century print advertising look, 1950s American graphic design, hand-painted script paired with condensed serif capitals, halftone photo collages cut out on flat color fields, restrained ornament from stars and thin rules, crisp legible small text, no gradients, no neon, no modern flat-UI look, no watermark, no checkerboard.",
137 2048,
138 2048,
139 12,
140 43,
141 "ming_image_0.1_design_int8_convrot.safetensors",
142 "ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors",
143 "ming_image_vae_bf16.safetensors",
144 true,
145 "qwen3.8_27b_w4a8.safetensors",
146 "comfy kitchen attention"
147 ],
148 "widgets_values_named": {
149 "value_2": "A retro 1950s mid-century e-commerce homepage design, full-page web layout, square canvas, vintage printed-paper aesthetic: aged cream background with fine halftone grain, slightly yellowed edges, thin dark-green hairline rules separating the sections, muted palette of deep pine green, teal, coral red-orange, warm red and off-white.\n\nTop: a full-width muted teal announcement bar with centered small uppercase serif text \"★ FREE SHIPPING ON ORDERS OVER $75 ★\", with a tiny star at each end. Below it the header, laid out on one row: on the left a horizontal menu in small dark-green uppercase condensed type reading \"NEW ARRIVALS WOMEN MEN ACCESSORIES SALE ABOUT\"; centered, the brand block with the tiny line \"EST. ★ 2010\" above a large warm-red hand-painted script wordmark \"Timeless\" with a thin off-white outline, then \"VINTAGE\" in bold dark-green serif capitals beneath it and a small \"CLOTHING CO.\" line with flanking rules; on the right three small dark-green line icons with tiny uppercase labels \"SEARCH\", \"ACCOUNT\", \"CART\", the cart icon in coral red. A horizontal hairline rule runs under the header.\n\nHero: a large rounded rectangle panel in pale mint with a thin dark-green border, containing on the left the small uppercase line \"★ AUTHENTIC STYLES. TIMELESS YOU.\", then the word \"Vintage\" in enormous warm-red hand-painted script with a soft off-white outline and a subtle extruded shadow, then \"NEVER FADES\" in heavy dark-green serif capitals, then two lines of centered grey-green body copy \"Handpicked vintage clothing from every era. Unique pieces with a story to tell.\", then a coral ribbon-shaped button with a notched tail and small star at each end containing the uppercase text \"SHOP NEW ARRIVALS →\". On the right side of the panel: a 1950s woman in a pale yellow cardigan, plaid full skirt, sunglasses and a polka-dot headscarf standing beside a pink 1950s sedan, plus a leaning vintage roadside sign painted in red and dark green with the text \"Timeless\" in script and \"VINTAGE CLOTHING CO.\" in capitals, and a small teal starburst badge to the right reading \"SUSTAINABLE FASHION TIMELESS STYLE\" in tiny stacked uppercase type.\n\nMiddle: a centered section heading \"★ SHOP BY CATEGORY ★\" in bold dark-green uppercase with small coral stars. Beneath it a single row of five circular product medallions, each a vintage photograph inside a cream ring with a thin dark-green outline: a smiling woman in a pink dress, a man in a brown suit, a brown leather handbag, a pair of dark loafers, and a blue denim jacket. Under each medallion sits a coral ribbon banner with a notched tail holding the uppercase label \"★ DRESSES ★\", \"★ MEN ★\", \"★ ACCESSORIES ★\", \"★ SHOES ★\", \"★ OUTERWEAR ★\", each with a small green star below the ribbon.\n\nLower band: a wide pale mint panel with a thin border. Inside on the left, a woman in a green blouse playing a vinyl record on a cream and brown suitcase-style record player. On the right, a warm photograph of a small vintage storefront with a striped awning and signs reading \"Timeless\" and \"VINTAGE FINDS\". Centered between them: \"More Than Clothes,\" in dark-green hand-written script, then \"★ IT'S A LIFESTYLE ★\" in bold dark-green uppercase, then three lines of grey-green body copy \"We believe in quality, individuality, and the charm of the past. Every piece is curated with care and meant to be loved for years to come.\", then a coral ribbon button reading \"READ OUR STORY →\".\n\nFooter: a darker teal band across the bottom with a small envelope icon and the heading \"JOIN OUR MAILING LIST\" beside the line \"Get updates on new arrivals and special offers!\", three small circular social icons in the middle (facebook, instagram, pinterest) under a thin vertical hairline, a heading \"CUSTOMER CARE\" with the items \"Shipping\", \"Returns\", \"FAQ\", \"Contact Us\", and on the right \"© 2024 TIMELESS VINTAGE\" with \"All rights reserved.\" beneath it.\n\nStyle notes: authentic mid-century print advertising look, 1950s American graphic design, hand-painted script paired with condensed serif capitals, halftone photo collages cut out on flat color fields, restrained ornament from stars and thin rules, crisp legible small text, no gradients, no neon, no modern flat-UI look, no watermark, no checkerboard.",
150 "value": 2048,
151 "value_1": 2048,
152 "steps": 12,
153 "noise_seed": 43,
154 "unet_name": "ming_image_0.1_design_int8_convrot.safetensors",
155 "clip_name": "ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors",
156 "vae_name": "ming_image_vae_bf16.safetensors",
157 "switch": true,
158 "clip_name_1": "qwen3.8_27b_w4a8.safetensors",
159 "attention": "comfy kitchen attention"
160 }
161 },
162 {
163 "id": 82,
164 "type": "MarkdownNote",
165 "pos": [
166 -2190,
167 3850
168 ],
169 "size": [
170 540,
171 670
172 ],
173 "flags": [],
174 "order": 1,
175 "mode": 0,
176 "inputs": [],
177 "outputs": [],
178 "properties": [],
179 "widgets_values": [
180 "Guide: [Subgraph](https://docs.comfy.org/interface/features/subgraph)\n\n## Model Links\n- [Hugging Face:Comfy-Org/Ming-Image](https://huggingface.co/Comfy-Org/Ming-Image)\n- [ModelScope:Comfy-Org/Ming-Image](https://modelscope.cn/models/Comfy-Org/Ming-Image)\n- [Hugging Face:Comfy-Org/Qwen3.8-27B](https://huggingface.co/Comfy-Org/Qwen3.8-27B)\n- [ModelScope:Comfy-Org/Qwen3.8-27B](https://modelscope.cn/models/Comfy-Org/Qwen3.8-27B)\n\n\n**text_encoders**\n\n- [qwen3.8_27b_w4a8.safetensors](https://huggingface.co/Comfy-Org/Qwen3.8-27B/resolve/main/text_encoders/qwen3.8_27b_w4a8.safetensors) (16.1 GB)\n- [ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/text_encoders/ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors) (18.17 GB)\n\n**diffusion_models**\n\n- [ming_image_0.1_design_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/diffusion_models/ming_image_0.1_design_int8_convrot.safetensors) (5.75 GB)\n\n**vae**\n\n- [ming_image_vae_bf16.safetensors](https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/vae/ming_image_vae_bf16.safetensors) (242.1 MB)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ ├── 📂 text_encoders/\n│ │ ├── qwen3.8_27b_w4a8.safetensors\n│ │ └── ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors\n│ ├── 📂 diffusion_models/\n│ │ └── ming_image_0.1_design_int8_convrot.safetensors\n│ └── 📂 vae/\n│ └── ming_image_vae_bf16.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n"
181 ],
182 "widgets_values_named": {
183 "text": "Guide: [Subgraph](https://docs.comfy.org/interface/features/subgraph)\n\n## Model Links\n- [Hugging Face:Comfy-Org/Ming-Image](https://huggingface.co/Comfy-Org/Ming-Image)\n- [ModelScope:Comfy-Org/Ming-Image](https://modelscope.cn/models/Comfy-Org/Ming-Image)\n- [Hugging Face:Comfy-Org/Qwen3.8-27B](https://huggingface.co/Comfy-Org/Qwen3.8-27B)\n- [ModelScope:Comfy-Org/Qwen3.8-27B](https://modelscope.cn/models/Comfy-Org/Qwen3.8-27B)\n\n\n**text_encoders**\n\n- [qwen3.8_27b_w4a8.safetensors](https://huggingface.co/Comfy-Org/Qwen3.8-27B/resolve/main/text_encoders/qwen3.8_27b_w4a8.safetensors) (16.1 GB)\n- [ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/text_encoders/ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors) (18.17 GB)\n\n**diffusion_models**\n\n- [ming_image_0.1_design_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/diffusion_models/ming_image_0.1_design_int8_convrot.safetensors) (5.75 GB)\n\n**vae**\n\n- [ming_image_vae_bf16.safetensors](https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/vae/ming_image_vae_bf16.safetensors) (242.1 MB)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ ├── 📂 text_encoders/\n│ │ ├── qwen3.8_27b_w4a8.safetensors\n│ │ └── ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors\n│ ├── 📂 diffusion_models/\n│ │ └── ming_image_0.1_design_int8_convrot.safetensors\n│ └── 📂 vae/\n│ └── ming_image_vae_bf16.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n"
184 },
185 "color": "#222",
186 "bgcolor": "#000"
187 },
188 {
189 "id": 83,
190 "type": "MarkdownNote",
191 "pos": [
192 -1620,
193 3850
194 ],
195 "size": [
196 540,
197 670
198 ],
199 "flags": [],
200 "order": 2,
201 "mode": 0,
202 "inputs": [],
203 "outputs": [],
204 "title": "Note: Ming Image 0.1 Design",
205 "properties": [],
206 "widgets_values": [
207 "## Ming Image 0.1 Design\n\nMing Image 0.1 Design is a 6B text-to-image model from inclusionAI (Ant Group), built for design work: UI screens, infographics, posters and other text-heavy layouts. It outputs RGBA images with a real alpha channel, so a result can be dropped straight into a design file. Released under the MIT license.\n\nWhat it is good at:\n\n- Long, structured layouts: pages, posters and infographics with many sections, written section by section.\n- Text rendering: strings quoted in the prompt are followed closely, so labels, prices and headings stay legible.\n- Transparent output: transparent backgrounds and isolated elements for compositing.\n- Prompt rewriting: an instruction-following VLM turns a short idea into a detailed layout caption before sampling.\n\nThe checkpoints this workflow needs, and where to put them, are listed in the **Model Links** note next to this one.\n\n## How to use this workflow\n\n1. Enter the **Text to Image (Ming Image 0.1 Design)** subgraph with the **Enter subgraph** button at the bottom of the node.\n2. Write your prompt in the `prompt` field, describing the layout section by section and putting any copy you want rendered in quotes.\n3. Choose the canvas with the `ResolutionSelector` node: 2048 x 2048 gives the best quality, 1024 x 1024 is faster.\n4. Keep the prompt enhancer on to let Qwen3.8-27B expand your prompt, or switch `enable_PE` off to send your text exactly as typed. The enhancer runs the rewriter system prompt released by the model authors, see below.\n5. Queue the prompt: the decoded image appears in the `PreviewImage` node.\n6. For a transparent background, prepend one RGBA phrase to your prompt, for example `RGBA, 4-channel, transparent background`.\n\n## Prompt guide\n\n- [Text-to-image prompt rewriting](https://github.com/inclusionAI/Ming-Image#text-to-image-prompt-rewriting): the official prompt structure and the [rewriter system prompt](https://github.com/inclusionAI/Ming-Image/blob/main/assets/t2i_rewriter_system_prompt.txt) that this workflow's prompt enhancer runs.\n- [Transparent-background generation tip](https://github.com/inclusionAI/Ming-Image#transparent-background-generation-tip): the fixed RGBA phrases to prepend.\n"
208 ],
209 "widgets_values_named": {
210 "text": "## Ming Image 0.1 Design\n\nMing Image 0.1 Design is a 6B text-to-image model from inclusionAI (Ant Group), built for design work: UI screens, infographics, posters and other text-heavy layouts. It outputs RGBA images with a real alpha channel, so a result can be dropped straight into a design file. Released under the MIT license.\n\nWhat it is good at:\n\n- Long, structured layouts: pages, posters and infographics with many sections, written section by section.\n- Text rendering: strings quoted in the prompt are followed closely, so labels, prices and headings stay legible.\n- Transparent output: transparent backgrounds and isolated elements for compositing.\n- Prompt rewriting: an instruction-following VLM turns a short idea into a detailed layout caption before sampling.\n\nThe checkpoints this workflow needs, and where to put them, are listed in the **Model Links** note next to this one.\n\n## How to use this workflow\n\n1. Enter the **Text to Image (Ming Image 0.1 Design)** subgraph with the **Enter subgraph** button at the bottom of the node.\n2. Write your prompt in the `prompt` field, describing the layout section by section and putting any copy you want rendered in quotes.\n3. Choose the canvas with the `ResolutionSelector` node: 2048 x 2048 gives the best quality, 1024 x 1024 is faster.\n4. Keep the prompt enhancer on to let Qwen3.8-27B expand your prompt, or switch `enable_PE` off to send your text exactly as typed. The enhancer runs the rewriter system prompt released by the model authors, see below.\n5. Queue the prompt: the decoded image appears in the `PreviewImage` node.\n6. For a transparent background, prepend one RGBA phrase to your prompt, for example `RGBA, 4-channel, transparent background`.\n\n## Prompt guide\n\n- [Text-to-image prompt rewriting](https://github.com/inclusionAI/Ming-Image#text-to-image-prompt-rewriting): the official prompt structure and the [rewriter system prompt](https://github.com/inclusionAI/Ming-Image/blob/main/assets/t2i_rewriter_system_prompt.txt) that this workflow's prompt enhancer runs.\n- [Transparent-background generation tip](https://github.com/inclusionAI/Ming-Image#transparent-background-generation-tip): the fixed RGBA phrases to prepend.\n"
211 },
212 "color": "#222",
213 "bgcolor": "#000"
214 },
215 {
216 "id": 84,
217 "type": "SaveImageAdvanced",
218 "pos": [
219 -210,
220 3860
221 ],
222 "size": [
223 720,
224 580
225 ],
226 "flags": [],
227 "order": 4,
228 "mode": 0,
229 "inputs": [
230 {
231 "name": "images",
232 "type": "IMAGE",
233 "link": 176
234 }
235 ],
236 "outputs": [
237 {
238 "name": "images",
239 "type": "IMAGE",
240 "links": null
241 }
242 ],
243 "properties": [],
244 "widgets_values": [
245 "ComfyUI",
246 "png",
247 "8-bit",
248 "sRGB"
249 ],
250 "widgets_values_named": {
251 "filename_prefix": "ComfyUI",
252 "format": "png",
253 "format.bit_depth": "8-bit",
254 "format.input_color_space": "sRGB"
255 }
256 }
257 ],
258 "links": [
259 [
260 163,
261 75,
262 0,
263 80,
264 1,
265 "INT"
266 ],
267 [
268 164,
269 75,
270 1,
271 80,
272 2,
273 "INT"
274 ],
275 [
276 176,
277 80,
278 0,
279 84,
280 0,
281 "IMAGE"
282 ]
283 ],
284 "groups": [],
285 "definitions": {
286 "subgraphs": [
287 {
288 "id": "4d166aa4-f08c-43b7-b889-2b43e3b4c535",
289 "version": 1,
290 "state": {
291 "lastGroupId": 5,
292 "lastNodeId": 85,
293 "lastLinkId": 180,
294 "lastRerouteId": 0
295 },
296 "revision": 0,
297 "config": [],
298 "name": "Text to Image (Ming Image 0.1 Design)",
299 "inputNode": {
300 "id": -10,
301 "bounding": [
302 -2880,
303 3820,
304 152.361328125,
305 268
306 ]
307 },
308 "outputNode": {
309 "id": -20,
310 "bounding": [
311 1910,
312 3830,
313 128,
314 68
315 ]
316 },
317 "inputs": [
318 {
319 "id": "88ae0295-52ee-4535-a1b9-bc30b15e2ad6",
320 "name": "value_2",
321 "type": "STRING",
322 "linkIds": [
323 166
324 ],
325 "label": "prompt",
326 "pos": [
327 -2751.638671875,
328 3844
329 ]
330 },
331 {
332 "id": "9fad0561-5db4-4e53-8ec1-248a285bdc46",
333 "name": "value",
334 "type": "INT",
335 "linkIds": [
336 154
337 ],
338 "localized_name": "value",
339 "label": "width",
340 "pos": [
341 -2751.638671875,
342 3864
343 ]
344 },
345 {
346 "id": "43c27ee0-be52-47a5-94a9-815281cb4653",
347 "name": "value_1",
348 "type": "INT",
349 "linkIds": [
350 155
351 ],
352 "localized_name": "value_1",
353 "label": "height",
354 "pos": [
355 -2751.638671875,
356 3884
357 ]
358 },
359 {
360 "id": "92e3c2dd-2c49-4e7c-aacb-7c9c0abcb613",
361 "name": "steps",
362 "type": "INT",
363 "linkIds": [
364 174
365 ],
366 "pos": [
367 -2751.638671875,
368 3904
369 ]
370 },
371 {
372 "id": "ef39e108-acee-4a3f-81b7-4f2392e832ea",
373 "name": "noise_seed",
374 "type": "INT",
375 "linkIds": [
376 173
377 ],
378 "pos": [
379 -2751.638671875,
380 3924
381 ]
382 },
383 {
384 "id": "cf84da08-33e4-4789-b104-e678f58bb83e",
385 "name": "unet_name",
386 "type": "COMBO",
387 "linkIds": [
388 172
389 ],
390 "pos": [
391 -2751.638671875,
392 3944
393 ]
394 },
395 {
396 "id": "ccdca2eb-b0cd-4f45-a6f1-ea05d9cb9273",
397 "name": "clip_name",
398 "type": "COMBO",
399 "linkIds": [
400 170
401 ],
402 "pos": [
403 -2751.638671875,
404 3964
405 ]
406 },
407 {
408 "id": "3ae49b9d-b2c4-471e-a89d-60f0a5effb61",
409 "name": "vae_name",
410 "type": "COMBO",
411 "linkIds": [
412 171
413 ],
414 "pos": [
415 -2751.638671875,
416 3984
417 ]
418 },
419 {
420 "id": "125f50ed-f30e-490f-8d92-8138c884aa78",
421 "name": "switch",
422 "type": "BOOLEAN",
423 "linkIds": [
424 167
425 ],
426 "label": "enable_PE",
427 "pos": [
428 -2751.638671875,
429 4004
430 ]
431 },
432 {
433 "id": "d0eebc8c-dd31-46d8-b316-e29c69d9388c",
434 "name": "clip_name_1",
435 "type": "COMBO",
436 "linkIds": [
437 175
438 ],
439 "label": "PE_model",
440 "pos": [
441 -2751.638671875,
442 4024
443 ]
444 },
445 {
446 "id": "96dbd402-e58d-4956-873b-51107ae8b4d8",
447 "name": "attention",
448 "type": "COMBO",
449 "linkIds": [
450 168
451 ],
452 "label": "attention_backend",
453 "pos": [
454 -2751.638671875,
455 4044
456 ]
457 }
458 ],
459 "outputs": [
460 {
461 "id": "559f4837-0984-4c6a-ad73-c8e8526139b2",
462 "name": "IMAGE",
463 "type": "IMAGE",
464 "linkIds": [
465 127
466 ],
467 "localized_name": "IMAGE",
468 "pos": [
469 1934,
470 3854
471 ]
472 }
473 ],
474 "widgets": [],
475 "nodes": [
476 {
477 "id": 46,
478 "type": "CLIPLoader",
479 "pos": [
480 -1740,
481 2950
482 ],
483 "size": [
484 490,
485 120
486 ],
487 "flags": [],
488 "order": 12,
489 "mode": 0,
490 "inputs": [
491 {
492 "localized_name": "clip_name",
493 "name": "clip_name",
494 "type": "COMBO",
495 "widget": {
496 "name": "clip_name"
497 },
498 "link": 175
499 }
500 ],
501 "outputs": [
502 {
503 "localized_name": "CLIP",
504 "name": "CLIP",
505 "type": "CLIP",
506 "links": [
507 178
508 ]
509 }
510 ],
511 "properties": {
512 "Node name for S&R": "CLIPLoader",
513 "models": [
514 {
515 "name": "qwen3.8_27b_w4a8.safetensors",
516 "url": "https://huggingface.co/Comfy-Org/Qwen3.8-27B/resolve/main/text_encoders/qwen3.8_27b_w4a8.safetensors",
517 "directory": "text_encoders"
518 }
519 ]
520 },
521 "widgets_values": [
522 "qwen3.8_27b_w4a8.safetensors",
523 "stable_diffusion",
524 "default"
525 ],
526 "widgets_values_named": {
527 "clip_name": "qwen3.8_27b_w4a8.safetensors",
528 "type": "stable_diffusion",
529 "device": "default"
530 },
531 "color": "#432",
532 "bgcolor": "#653"
533 },
534 {
535 "id": 45,
536 "type": "PrimitiveStringMultiline",
537 "pos": [
538 -1740,
539 3120
540 ],
541 "size": [
542 500,
543 760
544 ],
545 "flags": {
546 "collapsed": true
547 },
548 "order": 0,
549 "mode": 0,
550 "inputs": [],
551 "outputs": [
552 {
553 "localized_name": "STRING",
554 "name": "STRING",
555 "type": "STRING",
556 "links": [
557 179
558 ]
559 }
560 ],
561 "title": "Text (System Prompt)",
562 "properties": {
563 "Node name for S&R": "PrimitiveStringMultiline"
564 },
565 "widgets_values": [
566 "You are a senior visual designer and image-prompt engineer. Expand the user's request into one precise, high-resolution Figma-style caption. Return only one JSON object.\n\nUse exactly two top-level keys. `canvas_settings` contains exactly `aspect_ratio`, `ambient_lighting`, and `image_style`. `layers` lists visible groups from background to topmost overlay. Every layer contains exactly `description`, `coordinates`, `hierarchy_and_relation`, and `color_specs`; `color_specs` is an array of hex colors.\n\n`coordinates` MUST be one string, never an object or array, in exactly this form: `\"cx: 0.500, cy: 0.500, w: 1.000, h: 1.000\"`. Values are normalized; each bbox encloses its complete owned object and stays inside the canvas.\n\nA layer is one selectable visible semantic group: background, full person, coherent object, panel, card, row, or text block. Prefer the fewest groups that preserve the layout. Keep people and objects intact. Never create invisible parents, guides, placeholders, empty layers, duplicate summaries, or multiple owners for one element.\n\nPreserve every user-supplied rendered string character-for-character and as one contiguous string. Unless multiple visible copies are requested, it must occur exactly once across all `description` fields and zero times in `hierarchy_and_relation`. Quote it only where describing its visible rendering; refer to the related subject elsewhere with unquoted semantic wording. Enumerate intended copy, invent extra copy sparingly, and never hide content behind \"other text\", \"remaining labels\", or \"etc.\"\n\nDescribe concrete composition, typography, materials, texture, lighting, pose, and camera treatment without literary filler. Use `hierarchy_and_relation` only for ownership, alignment, containment, stacking, and occlusion.\n\nInfer structured layouts first. Use one complete layer per card and state its row and column. A compact secondary table may be one layer only if every header and cell is listed; otherwise use a visible shared frame when present, one complete header, and one complete layer per body row, binding values to columns and stating blanks. Enumerate sequences, schedules, spans, gaps, and vacant tracks in visual order. Do not mistake ordinary alignment for a table.\n\nSilently verify schema, string coordinates, Z-order, exact-text counts, geometry, bbox validity, and completeness."
567 ],
568 "widgets_values_named": {
569 "value": "You are a senior visual designer and image-prompt engineer. Expand the user's request into one precise, high-resolution Figma-style caption. Return only one JSON object.\n\nUse exactly two top-level keys. `canvas_settings` contains exactly `aspect_ratio`, `ambient_lighting`, and `image_style`. `layers` lists visible groups from background to topmost overlay. Every layer contains exactly `description`, `coordinates`, `hierarchy_and_relation`, and `color_specs`; `color_specs` is an array of hex colors.\n\n`coordinates` MUST be one string, never an object or array, in exactly this form: `\"cx: 0.500, cy: 0.500, w: 1.000, h: 1.000\"`. Values are normalized; each bbox encloses its complete owned object and stays inside the canvas.\n\nA layer is one selectable visible semantic group: background, full person, coherent object, panel, card, row, or text block. Prefer the fewest groups that preserve the layout. Keep people and objects intact. Never create invisible parents, guides, placeholders, empty layers, duplicate summaries, or multiple owners for one element.\n\nPreserve every user-supplied rendered string character-for-character and as one contiguous string. Unless multiple visible copies are requested, it must occur exactly once across all `description` fields and zero times in `hierarchy_and_relation`. Quote it only where describing its visible rendering; refer to the related subject elsewhere with unquoted semantic wording. Enumerate intended copy, invent extra copy sparingly, and never hide content behind \"other text\", \"remaining labels\", or \"etc.\"\n\nDescribe concrete composition, typography, materials, texture, lighting, pose, and camera treatment without literary filler. Use `hierarchy_and_relation` only for ownership, alignment, containment, stacking, and occlusion.\n\nInfer structured layouts first. Use one complete layer per card and state its row and column. A compact secondary table may be one layer only if every header and cell is listed; otherwise use a visible shared frame when present, one complete header, and one complete layer per body row, binding values to columns and stating blanks. Enumerate sequences, schedules, spans, gaps, and vacant tracks in visual order. Do not mistake ordinary alignment for a table.\n\nSilently verify schema, string coordinates, Z-order, exact-text counts, geometry, bbox validity, and completeness."
570 }
571 },
572 {
573 "id": 77,
574 "type": "PrimitiveStringMultiline",
575 "pos": [
576 -1730,
577 3200
578 ],
579 "size": [
580 500,
581 520
582 ],
583 "flags": [],
584 "order": 18,
585 "mode": 0,
586 "inputs": [
587 {
588 "localized_name": "value",
589 "name": "value",
590 "type": "STRING",
591 "widget": {
592 "name": "value"
593 },
594 "link": 166
595 }
596 ],
597 "outputs": [
598 {
599 "localized_name": "STRING",
600 "name": "STRING",
601 "type": "STRING",
602 "links": [
603 158,
604 180
605 ]
606 }
607 ],
608 "title": "Text (User Input)",
609 "properties": {
610 "Node name for S&R": "PrimitiveStringMultiline"
611 },
612 "widgets_values": [
613 ""
614 ],
615 "widgets_values_named": {
616 "value": ""
617 }
618 },
619 {
620 "id": 40,
621 "type": "CLIPTextEncode",
622 "pos": [
623 -1090,
624 3940
625 ],
626 "size": [
627 490,
628 520
629 ],
630 "flags": {
631 "collapsed": false
632 },
633 "order": 9,
634 "mode": 0,
635 "inputs": [
636 {
637 "localized_name": "clip",
638 "name": "clip",
639 "type": "CLIP",
640 "link": 72
641 },
642 {
643 "localized_name": "text",
644 "name": "text",
645 "type": "STRING",
646 "widget": {
647 "name": "text"
648 },
649 "link": 162
650 }
651 ],
652 "outputs": [
653 {
654 "localized_name": "CONDITIONING",
655 "name": "CONDITIONING",
656 "type": "CONDITIONING",
657 "links": [
658 152
659 ]
660 }
661 ],
662 "properties": {
663 "Node name for S&R": "CLIPTextEncode"
664 },
665 "widgets_values": [
666 ""
667 ],
668 "widgets_values_named": {
669 "text": ""
670 },
671 "color": "#232",
672 "bgcolor": "#353"
673 },
674 {
675 "id": 23,
676 "type": "CLIPLoader",
677 "pos": [
678 -1720,
679 4140
680 ],
681 "size": [
682 510,
683 120
684 ],
685 "flags": [],
686 "order": 7,
687 "mode": 0,
688 "inputs": [
689 {
690 "localized_name": "clip_name",
691 "name": "clip_name",
692 "type": "COMBO",
693 "widget": {
694 "name": "clip_name"
695 },
696 "link": 170
697 }
698 ],
699 "outputs": [
700 {
701 "localized_name": "CLIP",
702 "name": "CLIP",
703 "type": "CLIP",
704 "links": [
705 72
706 ]
707 }
708 ],
709 "properties": {
710 "Node name for S&R": "CLIPLoader",
711 "models": [
712 {
713 "name": "ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors",
714 "url": "https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/text_encoders/ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors",
715 "directory": "text_encoders"
716 }
717 ]
718 },
719 "widgets_values": [
720 "ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors",
721 "qwen_image",
722 "default"
723 ],
724 "widgets_values_named": {
725 "clip_name": "ming_image_0.1_ling_mini_2.0_int8_convrot.safetensors",
726 "type": "qwen_image",
727 "device": "default"
728 },
729 "color": "#432",
730 "bgcolor": "#653"
731 },
732 {
733 "id": 1,
734 "type": "UNETLoader",
735 "pos": [
736 -1720,
737 3980
738 ],
739 "size": [
740 510,
741 90
742 ],
743 "flags": [],
744 "order": 2,
745 "mode": 0,
746 "inputs": [
747 {
748 "localized_name": "unet_name",
749 "name": "unet_name",
750 "type": "COMBO",
751 "widget": {
752 "name": "unet_name"
753 },
754 "link": 172
755 }
756 ],
757 "outputs": [
758 {
759 "localized_name": "MODEL",
760 "name": "MODEL",
761 "type": "MODEL",
762 "links": [
763 142
764 ]
765 }
766 ],
767 "properties": {
768 "Node name for S&R": "UNETLoader",
769 "models": [
770 {
771 "name": "ming_image_0.1_design_int8_convrot.safetensors",
772 "url": "https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/diffusion_models/ming_image_0.1_design_int8_convrot.safetensors",
773 "directory": "diffusion_models"
774 }
775 ]
776 },
777 "widgets_values": [
778 "ming_image_0.1_design_int8_convrot.safetensors",
779 "default"
780 ],
781 "widgets_values_named": {
782 "unet_name": "ming_image_0.1_design_int8_convrot.safetensors",
783 "weight_dtype": "default"
784 },
785 "color": "#223",
786 "bgcolor": "#335"
787 },
788 {
789 "id": 42,
790 "type": "ModelSamplingFlux",
791 "pos": [
792 -540,
793 4320
794 ],
795 "size": [
796 270,
797 130
798 ],
799 "flags": {
800 "collapsed": false
801 },
802 "order": 11,
803 "mode": 0,
804 "inputs": [
805 {
806 "localized_name": "model",
807 "name": "model",
808 "type": "MODEL",
809 "link": 143
810 },
811 {
812 "localized_name": "width",
813 "name": "width",
814 "type": "INT",
815 "widget": {
816 "name": "width"
817 },
818 "link": 128
819 },
820 {
821 "localized_name": "height",
822 "name": "height",
823 "type": "INT",
824 "widget": {
825 "name": "height"
826 },
827 "link": 131
828 }
829 ],
830 "outputs": [
831 {
832 "localized_name": "MODEL",
833 "name": "MODEL",
834 "type": "MODEL",
835 "links": [
836 81,
837 150
838 ]
839 }
840 ],
841 "properties": {
842 "Node name for S&R": "ModelSamplingFlux"
843 },
844 "widgets_values": [
845 1.15,
846 0.5,
847 1536,
848 1536
849 ],
850 "widgets_values_named": {
851 "max_shift": 1.15,
852 "base_shift": 0.5,
853 "width": 1536,
854 "height": 1536
855 }
856 },
857 {
858 "id": 68,
859 "type": "ModelAttentionBackend",
860 "pos": [
861 -540,
862 4170
863 ],
864 "size": [
865 270,
866 60
867 ],
868 "flags": [],
869 "order": 15,
870 "mode": 0,
871 "inputs": [
872 {
873 "localized_name": "model",
874 "name": "model",
875 "type": "MODEL",
876 "link": 142
877 },
878 {
879 "localized_name": "backend",
880 "name": "attention",
881 "type": "COMBO",
882 "widget": {
883 "name": "attention"
884 },
885 "link": 168
886 }
887 ],
888 "outputs": [
889 {
890 "localized_name": "model",
891 "name": "model",
892 "type": "MODEL",
893 "links": [
894 143
895 ]
896 }
897 ],
898 "properties": {
899 "Node name for S&R": "ModelAttentionBackend"
900 },
901 "widgets_values": [
902 "comfy kitchen attention"
903 ],
904 "widgets_values_named": {
905 "attention": "comfy kitchen attention"
906 }
907 },
908 {
909 "id": 76,
910 "type": "ComfySwitchNode",
911 "pos": [
912 -620,
913 2960
914 ],
915 "size": [
916 270,
917 80
918 ],
919 "flags": [],
920 "order": 17,
921 "mode": 0,
922 "inputs": [
923 {
924 "localized_name": "on_false",
925 "name": "on_false",
926 "shape": 7,
927 "type": "STRING",
928 "link": 158
929 },
930 {
931 "localized_name": "on_true",
932 "name": "on_true",
933 "shape": 7,
934 "type": "STRING",
935 "link": 177
936 },
937 {
938 "localized_name": "switch",
939 "name": "switch",
940 "type": "BOOLEAN",
941 "widget": {
942 "name": "switch"
943 },
944 "link": 167
945 }
946 ],
947 "outputs": [
948 {
949 "localized_name": "output",
950 "name": "output",
951 "type": "STRING",
952 "links": [
953 161
954 ]
955 }
956 ],
957 "properties": {
958 "Node name for S&R": "ComfySwitchNode"
959 },
960 "widgets_values": [
961 false
962 ],
963 "widgets_values_named": {
964 "switch": false
965 }
966 },
967 {
968 "id": 14,
969 "type": "VAELoader",
970 "pos": [
971 -1720,
972 4330
973 ],
974 "size": [
975 510,
976 70
977 ],
978 "flags": [],
979 "order": 4,
980 "mode": 0,
981 "inputs": [
982 {
983 "localized_name": "vae_name",
984 "name": "vae_name",
985 "type": "COMBO",
986 "widget": {
987 "name": "vae_name"
988 },
989 "link": 171
990 }
991 ],
992 "outputs": [
993 {
994 "localized_name": "VAE",
995 "name": "VAE",
996 "type": "VAE",
997 "links": [
998 21
999 ]
1000 }
1001 ],
1002 "properties": {
1003 "Node name for S&R": "VAELoader",
1004 "models": [
1005 {
1006 "name": "ming_image_vae_bf16.safetensors",
1007 "url": "https://huggingface.co/Comfy-Org/Ming-Image/resolve/main/vae/ming_image_vae_bf16.safetensors",
1008 "directory": "vae"
1009 }
1010 ]
1011 },
1012 "widgets_values": [
1013 "ming_image_vae_bf16.safetensors"
1014 ],
1015 "widgets_values_named": {
1016 "vae_name": "ming_image_vae_bf16.safetensors"
1017 },
1018 "color": "#322",
1019 "bgcolor": "#533"
1020 },
1021 {
1022 "id": 21,
1023 "type": "BasicScheduler",
1024 "pos": [
1025 -210,
1026 4360
1027 ],
1028 "size": [
1029 270,
1030 120
1031 ],
1032 "flags": [],
1033 "order": 6,
1034 "mode": 0,
1035 "inputs": [
1036 {
1037 "localized_name": "model",
1038 "name": "model",
1039 "type": "MODEL",
1040 "link": 81
1041 },
1042 {
1043 "localized_name": "steps",
1044 "name": "steps",
1045 "type": "INT",
1046 "widget": {
1047 "name": "steps"
1048 },
1049 "link": 174
1050 }
1051 ],
1052 "outputs": [
1053 {
1054 "localized_name": "SIGMAS",
1055 "name": "SIGMAS",
1056 "type": "SIGMAS",
1057 "links": [
1058 67
1059 ]
1060 }
1061 ],
1062 "properties": {
1063 "Node name for S&R": "BasicScheduler"
1064 },
1065 "widgets_values": [
1066 "simple",
1067 12,
1068 1
1069 ],
1070 "widgets_values_named": {
1071 "scheduler": "simple",
1072 "steps": 12,
1073 "denoise": 1
1074 }
1075 },
1076 {
1077 "id": 18,
1078 "type": "KSamplerSelect",
1079 "pos": [
1080 -210,
1081 4100
1082 ],
1083 "size": [
1084 270,
1085 70
1086 ],
1087 "flags": [],
1088 "order": 1,
1089 "mode": 0,
1090 "inputs": [],
1091 "outputs": [
1092 {
1093 "localized_name": "SAMPLER",
1094 "name": "SAMPLER",
1095 "type": "SAMPLER",
1096 "links": [
1097 25
1098 ]
1099 }
1100 ],
1101 "properties": {
1102 "Node name for S&R": "KSamplerSelect"
1103 },
1104 "widgets_values": [
1105 "euler"
1106 ],
1107 "widgets_values_named": {
1108 "sampler_name": "euler"
1109 }
1110 },
1111 {
1112 "id": 26,
1113 "type": "RandomNoise",
1114 "pos": [
1115 -210,
1116 3930
1117 ],
1118 "size": [
1119 270,
1120 90
1121 ],
1122 "flags": [],
1123 "order": 8,
1124 "mode": 0,
1125 "inputs": [
1126 {
1127 "localized_name": "noise_seed",
1128 "name": "noise_seed",
1129 "type": "INT",
1130 "widget": {
1131 "name": "noise_seed"
1132 },
1133 "link": 173
1134 }
1135 ],
1136 "outputs": [
1137 {
1138 "localized_name": "NOISE",
1139 "name": "NOISE",
1140 "type": "NOISE",
1141 "links": [
1142 38
1143 ]
1144 }
1145 ],
1146 "properties": {
1147 "Node name for S&R": "RandomNoise"
1148 },
1149 "widgets_values": [
1150 43,
1151 "randomize"
1152 ],
1153 "widgets_values_named": {
1154 "noise_seed": 43,
1155 "control_after_generate": "randomize"
1156 }
1157 },
1158 {
1159 "id": 73,
1160 "type": "BasicGuider",
1161 "pos": [
1162 -200,
1163 4240
1164 ],
1165 "size": [
1166 250,
1167 50
1168 ],
1169 "flags": [],
1170 "order": 16,
1171 "mode": 0,
1172 "inputs": [
1173 {
1174 "localized_name": "model",
1175 "name": "model",
1176 "type": "MODEL",
1177 "link": 150
1178 },
1179 {
1180 "localized_name": "conditioning",
1181 "name": "conditioning",
1182 "type": "CONDITIONING",
1183 "link": 152
1184 }
1185 ],
1186 "outputs": [
1187 {
1188 "localized_name": "GUIDER",
1189 "name": "GUIDER",
1190 "type": "GUIDER",
1191 "links": [
1192 151
1193 ]
1194 }
1195 ],
1196 "properties": {
1197 "Node name for S&R": "BasicGuider"
1198 }
1199 },
1200 {
1201 "id": 79,
1202 "type": "PreviewAny",
1203 "pos": [
1204 -230,
1205 2950
1206 ],
1207 "size": [
1208 640,
1209 790
1210 ],
1211 "flags": [],
1212 "order": 19,
1213 "mode": 0,
1214 "inputs": [
1215 {
1216 "localized_name": "source",
1217 "name": "source",
1218 "type": "*",
1219 "link": 161
1220 }
1221 ],
1222 "outputs": [
1223 {
1224 "localized_name": "STRING",
1225 "name": "STRING",
1226 "type": "STRING",
1227 "links": [
1228 162
1229 ]
1230 }
1231 ],
1232 "properties": {
1233 "Node name for S&R": "PreviewAny"
1234 },
1235 "widgets_values": [],
1236 "widgets_values_named": []
1237 },
1238 {
1239 "id": 41,
1240 "type": "EmptyLatentImage",
1241 "pos": [
1242 -1420,
1243 4650
1244 ],
1245 "size": [
1246 230,
1247 120
1248 ],
1249 "flags": {
1250 "collapsed": false
1251 },
1252 "order": 10,
1253 "mode": 0,
1254 "inputs": [
1255 {
1256 "localized_name": "width",
1257 "name": "width",
1258 "type": "INT",
1259 "widget": {
1260 "name": "width"
1261 },
1262 "link": 129
1263 },
1264 {
1265 "localized_name": "height",
1266 "name": "height",
1267 "type": "INT",
1268 "widget": {
1269 "name": "height"
1270 },
1271 "link": 130
1272 }
1273 ],
1274 "outputs": [
1275 {
1276 "localized_name": "LATENT",
1277 "name": "LATENT",
1278 "type": "LATENT",
1279 "links": [
1280 153
1281 ]
1282 }
1283 ],
1284 "properties": {
1285 "Node name for S&R": "EmptyLatentImage"
1286 },
1287 "widgets_values": [
1288 1024,
1289 2048,
1290 1
1291 ],
1292 "widgets_values_named": {
1293 "width": 1024,
1294 "height": 2048,
1295 "batch_size": 1
1296 }
1297 },
1298 {
1299 "id": 59,
1300 "type": "PrimitiveInt",
1301 "pos": [
1302 -1730,
1303 4590
1304 ],
1305 "size": [
1306 270,
1307 90
1308 ],
1309 "flags": [],
1310 "order": 13,
1311 "mode": 0,
1312 "inputs": [
1313 {
1314 "localized_name": "value",
1315 "name": "value",
1316 "type": "INT",
1317 "widget": {
1318 "name": "value"
1319 },
1320 "link": 154
1321 }
1322 ],
1323 "outputs": [
1324 {
1325 "localized_name": "INT",
1326 "name": "INT",
1327 "type": "INT",
1328 "links": [
1329 128,
1330 129
1331 ]
1332 }
1333 ],
1334 "title": "width",
1335 "properties": {
1336 "Node name for S&R": "PrimitiveInt"
1337 },
1338 "widgets_values": [
1339 2048,
1340 "fixed"
1341 ],
1342 "widgets_values_named": {
1343 "value": 2048,
1344 "control_after_generate": "fixed"
1345 }
1346 },
1347 {
1348 "id": 60,
1349 "type": "PrimitiveInt",
1350 "pos": [
1351 -1730,
1352 4730
1353 ],
1354 "size": [
1355 260,
1356 90
1357 ],
1358 "flags": [],
1359 "order": 14,
1360 "mode": 0,
1361 "inputs": [
1362 {
1363 "localized_name": "value",
1364 "name": "value",
1365 "type": "INT",
1366 "widget": {
1367 "name": "value"
1368 },
1369 "link": 155
1370 }
1371 ],
1372 "outputs": [
1373 {
1374 "localized_name": "INT",
1375 "name": "INT",
1376 "type": "INT",
1377 "links": [
1378 130,
1379 131
1380 ]
1381 }
1382 ],
1383 "title": "height",
1384 "properties": {
1385 "Node name for S&R": "PrimitiveInt"
1386 },
1387 "widgets_values": [
1388 2048,
1389 "fixed"
1390 ],
1391 "widgets_values_named": {
1392 "value": 2048,
1393 "control_after_generate": "fixed"
1394 }
1395 },
1396 {
1397 "id": 15,
1398 "type": "VAEDecode",
1399 "pos": [
1400 180,
1401 4580
1402 ],
1403 "size": [
1404 260,
1405 60
1406 ],
1407 "flags": [],
1408 "order": 5,
1409 "mode": 0,
1410 "inputs": [
1411 {
1412 "localized_name": "samples",
1413 "name": "samples",
1414 "type": "LATENT",
1415 "link": 149
1416 },
1417 {
1418 "localized_name": "vae",
1419 "name": "vae",
1420 "type": "VAE",
1421 "link": 21
1422 }
1423 ],
1424 "outputs": [
1425 {
1426 "localized_name": "IMAGE",
1427 "name": "IMAGE",
1428 "type": "IMAGE",
1429 "links": [
1430 127
1431 ]
1432 }
1433 ],
1434 "properties": {
1435 "Node name for S&R": "VAEDecode"
1436 },
1437 "color": "#322",
1438 "bgcolor": "#533"
1439 },
1440 {
1441 "id": 3,
1442 "type": "SamplerCustomAdvanced",
1443 "pos": [
1444 150,
1445 3930
1446 ],
1447 "size": [
1448 310,
1449 420
1450 ],
1451 "flags": [],
1452 "order": 3,
1453 "mode": 0,
1454 "inputs": [
1455 {
1456 "localized_name": "noise",
1457 "name": "noise",
1458 "type": "NOISE",
1459 "link": 38
1460 },
1461 {
1462 "localized_name": "guider",
1463 "name": "guider",
1464 "type": "GUIDER",
1465 "link": 151
1466 },
1467 {
1468 "localized_name": "sampler",
1469 "name": "sampler",
1470 "type": "SAMPLER",
1471 "link": 25
1472 },
1473 {
1474 "localized_name": "sigmas",
1475 "name": "sigmas",
1476 "type": "SIGMAS",
1477 "link": 67
1478 },
1479 {
1480 "localized_name": "latent_image",
1481 "name": "latent_image",
1482 "type": "LATENT",
1483 "link": 153
1484 }
1485 ],
1486 "outputs": [
1487 {
1488 "localized_name": "output",
1489 "name": "output",
1490 "type": "LATENT",
1491 "links": [
1492 149
1493 ]
1494 },
1495 {
1496 "localized_name": "denoised_output",
1497 "name": "denoised_output",
1498 "type": "LATENT",
1499 "links": null
1500 }
1501 ],
1502 "properties": {
1503 "Node name for S&R": "SamplerCustomAdvanced"
1504 }
1505 },
1506 {
1507 "id": 85,
1508 "type": "TextGenerateV2",
1509 "pos": [
1510 -1120,
1511 2970
1512 ],
1513 "size": [
1514 400,
1515 456
1516 ],
1517 "flags": [],
1518 "order": 20,
1519 "mode": 0,
1520 "showAdvanced": true,
1521 "inputs": [
1522 {
1523 "localized_name": "clip",
1524 "name": "clip",
1525 "type": "CLIP",
1526 "link": 178
1527 },
1528 {
1529 "localized_name": "image1",
1530 "name": "images.image1",
1531 "shape": 7,
1532 "type": "IMAGE",
1533 "link": null
1534 },
1535 {
1536 "localized_name": "video1",
1537 "name": "videos.video1",
1538 "shape": 7,
1539 "type": "IMAGE",
1540 "link": null
1541 },
1542 {
1543 "localized_name": "audio1",
1544 "name": "audios.audio1",
1545 "shape": 7,
1546 "type": "AUDIO",
1547 "link": null
1548 },
1549 {
1550 "localized_name": "system_prompt",
1551 "name": "system_prompt",
1552 "shape": 7,
1553 "type": "STRING",
1554 "link": 179
1555 },
1556 {
1557 "localized_name": "prompt",
1558 "name": "prompt",
1559 "type": "STRING",
1560 "widget": {
1561 "name": "prompt"
1562 },
1563 "link": 180
1564 }
1565 ],
1566 "outputs": [
1567 {
1568 "localized_name": "generated_text",
1569 "name": "generated_text",
1570 "type": "STRING",
1571 "links": [
1572 177
1573 ]
1574 },
1575 {
1576 "localized_name": "thinking",
1577 "name": "thinking",
1578 "type": "STRING",
1579 "links": null
1580 }
1581 ],
1582 "properties": {
1583 "Node name for S&R": "TextGenerateV2"
1584 },
1585 "widgets_values": [
1586 "",
1587 4096,
1588 "on",
1589 0.7,
1590 64,
1591 0.95,
1592 0.05,
1593 1.05,
1594 0,
1595 0,
1596 false,
1597 true,
1598 "auto"
1599 ],
1600 "widgets_values_named": {
1601 "prompt": "",
1602 "max_length": 4096,
1603 "sampling_mode": "on",
1604 "sampling_mode.temperature": 0.7,
1605 "sampling_mode.top_k": 64,
1606 "sampling_mode.top_p": 0.95,
1607 "sampling_mode.min_p": 0.05,
1608 "sampling_mode.repetition_penalty": 1.05,
1609 "sampling_mode.presence_penalty": 0,
1610 "sampling_mode.seed": 0,
1611 "thinking": false,
1612 "use_default_template": true,
1613 "mtp": "auto"
1614 }
1615 }
1616 ],
1617 "groups": [
1618 {
1619 "id": 2,
1620 "title": "Conditioning",
1621 "bounding": [
1622 -1120,
1623 3850,
1624 540,
1625 640
1626 ],
1627 "color": "#3f789e",
1628 "flags": []
1629 },
1630 {
1631 "id": 3,
1632 "title": "Models",
1633 "bounding": [
1634 -1750,
1635 3850,
1636 600,
1637 640
1638 ],
1639 "color": "#3f789e",
1640 "flags": []
1641 },
1642 {
1643 "id": 4,
1644 "title": "Size",
1645 "bounding": [
1646 -1750,
1647 4520,
1648 600,
1649 320
1650 ],
1651 "color": "#3f789e",
1652 "flags": []
1653 },
1654 {
1655 "id": 1,
1656 "title": "Sampling",
1657 "bounding": [
1658 -550,
1659 3850,
1660 1030,
1661 640
1662 ],
1663 "color": "#3f789e",
1664 "flags": []
1665 },
1666 {
1667 "id": 5,
1668 "title": "Prompt Enhancer",
1669 "bounding": [
1670 -1750,
1671 2880,
1672 2230,
1673 940
1674 ],
1675 "color": "#444",
1676 "flags": []
1677 }
1678 ],
1679 "links": [
1680 {
1681 "id": 72,
1682 "origin_id": 23,
1683 "origin_slot": 0,
1684 "target_id": 40,
1685 "target_slot": 0,
1686 "type": "CLIP"
1687 },
1688 {
1689 "id": 162,
1690 "origin_id": 79,
1691 "origin_slot": 0,
1692 "target_id": 40,
1693 "target_slot": 1,
1694 "type": "STRING"
1695 },
1696 {
1697 "id": 143,
1698 "origin_id": 68,
1699 "origin_slot": 0,
1700 "target_id": 42,
1701 "target_slot": 0,
1702 "type": "MODEL"
1703 },
1704 {
1705 "id": 142,
1706 "origin_id": 1,
1707 "origin_slot": 0,
1708 "target_id": 68,
1709 "target_slot": 0,
1710 "type": "MODEL"
1711 },
1712 {
1713 "id": 158,
1714 "origin_id": 77,
1715 "origin_slot": 0,
1716 "target_id": 76,
1717 "target_slot": 0,
1718 "type": "STRING"
1719 },
1720 {
1721 "id": 81,
1722 "origin_id": 42,
1723 "origin_slot": 0,
1724 "target_id": 21,
1725 "target_slot": 0,
1726 "type": "MODEL"
1727 },
1728 {
1729 "id": 150,
1730 "origin_id": 42,
1731 "origin_slot": 0,
1732 "target_id": 73,
1733 "target_slot": 0,
1734 "type": "MODEL"
1735 },
1736 {
1737 "id": 152,
1738 "origin_id": 40,
1739 "origin_slot": 0,
1740 "target_id": 73,
1741 "target_slot": 1,
1742 "type": "CONDITIONING"
1743 },
1744 {
1745 "id": 161,
1746 "origin_id": 76,
1747 "origin_slot": 0,
1748 "target_id": 79,
1749 "target_slot": 0,
1750 "type": "STRING"
1751 },
1752 {
1753 "id": 129,
1754 "origin_id": 59,
1755 "origin_slot": 0,
1756 "target_id": 41,
1757 "target_slot": 0,
1758 "type": "INT"
1759 },
1760 {
1761 "id": 130,
1762 "origin_id": 60,
1763 "origin_slot": 0,
1764 "target_id": 41,
1765 "target_slot": 1,
1766 "type": "INT"
1767 },
1768 {
1769 "id": 149,
1770 "origin_id": 3,
1771 "origin_slot": 0,
1772 "target_id": 15,
1773 "target_slot": 0,
1774 "type": "LATENT"
1775 },
1776 {
1777 "id": 21,
1778 "origin_id": 14,
1779 "origin_slot": 0,
1780 "target_id": 15,
1781 "target_slot": 1,
1782 "type": "VAE"
1783 },
1784 {
1785 "id": 38,
1786 "origin_id": 26,
1787 "origin_slot": 0,
1788 "target_id": 3,
1789 "target_slot": 0,
1790 "type": "NOISE"
1791 },
1792 {
1793 "id": 151,
1794 "origin_id": 73,
1795 "origin_slot": 0,
1796 "target_id": 3,
1797 "target_slot": 1,
1798 "type": "GUIDER"
1799 },
1800 {
1801 "id": 25,
1802 "origin_id": 18,
1803 "origin_slot": 0,
1804 "target_id": 3,
1805 "target_slot": 2,
1806 "type": "SAMPLER"
1807 },
1808 {
1809 "id": 67,
1810 "origin_id": 21,
1811 "origin_slot": 0,
1812 "target_id": 3,
1813 "target_slot": 3,
1814 "type": "SIGMAS"
1815 },
1816 {
1817 "id": 153,
1818 "origin_id": 41,
1819 "origin_slot": 0,
1820 "target_id": 3,
1821 "target_slot": 4,
1822 "type": "LATENT"
1823 },
1824 {
1825 "id": 127,
1826 "origin_id": 15,
1827 "origin_slot": 0,
1828 "target_id": -20,
1829 "target_slot": 0,
1830 "type": "IMAGE"
1831 },
1832 {
1833 "id": 166,
1834 "origin_id": -10,
1835 "origin_slot": 0,
1836 "target_id": 77,
1837 "target_slot": 0,
1838 "type": "STRING"
1839 },
1840 {
1841 "id": 154,
1842 "origin_id": -10,
1843 "origin_slot": 1,
1844 "target_id": 59,
1845 "target_slot": 0,
1846 "type": "INT"
1847 },
1848 {
1849 "id": 155,
1850 "origin_id": -10,
1851 "origin_slot": 2,
1852 "target_id": 60,
1853 "target_slot": 0,
1854 "type": "INT"
1855 },
1856 {
1857 "id": 173,
1858 "origin_id": -10,
1859 "origin_slot": 4,
1860 "target_id": 26,
1861 "target_slot": 0,
1862 "type": "INT"
1863 },
1864 {
1865 "id": 172,
1866 "origin_id": -10,
1867 "origin_slot": 5,
1868 "target_id": 1,
1869 "target_slot": 0,
1870 "type": "COMBO"
1871 },
1872 {
1873 "id": 170,
1874 "origin_id": -10,
1875 "origin_slot": 6,
1876 "target_id": 23,
1877 "target_slot": 0,
1878 "type": "COMBO"
1879 },
1880 {
1881 "id": 171,
1882 "origin_id": -10,
1883 "origin_slot": 7,
1884 "target_id": 14,
1885 "target_slot": 0,
1886 "type": "COMBO"
1887 },
1888 {
1889 "id": 167,
1890 "origin_id": -10,
1891 "origin_slot": 8,
1892 "target_id": 76,
1893 "target_slot": 2,
1894 "type": "BOOLEAN"
1895 },
1896 {
1897 "id": 175,
1898 "origin_id": -10,
1899 "origin_slot": 9,
1900 "target_id": 46,
1901 "target_slot": 0,
1902 "type": "COMBO"
1903 },
1904 {
1905 "id": 168,
1906 "origin_id": -10,
1907 "origin_slot": 10,
1908 "target_id": 68,
1909 "target_slot": 1,
1910 "type": "COMBO"
1911 },
1912 {
1913 "id": 128,
1914 "origin_id": 59,
1915 "origin_slot": 0,
1916 "target_id": 42,
1917 "target_slot": 1,
1918 "type": "INT"
1919 },
1920 {
1921 "id": 131,
1922 "origin_id": 60,
1923 "origin_slot": 0,
1924 "target_id": 42,
1925 "target_slot": 2,
1926 "type": "INT"
1927 },
1928 {
1929 "id": 174,
1930 "origin_id": -10,
1931 "origin_slot": 3,
1932 "target_id": 21,
1933 "target_slot": 1,
1934 "type": "INT"
1935 },
1936 {
1937 "id": 177,
1938 "origin_id": 85,
1939 "origin_slot": 0,
1940 "target_id": 76,
1941 "target_slot": 1,
1942 "type": "STRING"
1943 },
1944 {
1945 "id": 178,
1946 "origin_id": 46,
1947 "origin_slot": 0,
1948 "target_id": 85,
1949 "target_slot": 0,
1950 "type": "CLIP"
1951 },
1952 {
1953 "id": 179,
1954 "origin_id": 45,
1955 "origin_slot": 0,
1956 "target_id": 85,
1957 "target_slot": 4,
1958 "type": "STRING"
1959 },
1960 {
1961 "id": 180,
1962 "origin_id": 77,
1963 "origin_slot": 0,
1964 "target_id": 85,
1965 "target_slot": 5,
1966 "type": "STRING"
1967 }
1968 ],
1969 "extra": []
1970 }
1971 ]
1972 },
1973 "config": [],
1974 "extra": {
1975 "ds": {
1976 "scale": 0.7306341235399986,
1977 "offset": [
1978 1801.4394461122881,
1979 -3481.6098133117302
1980 ]
1981 },
1982 "frontendVersion": "1.57.6",
1983 "VHS_latentpreview": false,
1984 "VHS_latentpreviewrate": 0,
1985 "VHS_MetadataImage": true,
1986 "VHS_KeepIntermediate": true
1987 },
1988 "version": 0.4
1989}