ComfyPilotComfyPilot
工作流 / LLM / Qwen3-VL：文本生成
Qwen3-VL：文本生成

Qwen3-VL：文本生成

Qwen3-VL: Text Generation

使用Qwen3-VL模型，通过视觉理解生成文本，处理一张输入图像以产生上下文相关的文本响应。适用于图像描述、视觉问答和多模态内容生成。

LLM 5 个节点 1 个输入 129 次使用 ComfyUI ≥ 0.44.0 ComfyUI 8 KB

下载工作流 JSON 查看官方来源

下载后把文件拖进 ComfyUI 窗口即可打开。

需要的权重

这个工作流引用了 1 个权重文件，其中 1 个能在本站模型库里找到。

qwen3vl_4b_fp8_scaled.safetensors 在模型库查看

节点构成

CLIPLoader × 1
PreviewAny × 1
LoadImage × 1
MarkdownNote × 1
TextGenerateV2 × 1

官方标注的模型

Qwen 3.0

工作流 JSON

和上面「下载工作流 JSON」是同一份文件。

1{
2 "id": "11c31764-20a2-468e-846a-c4fa6abe0433",
3 "revision": 0,
4 "last_node_id": 10,
5 "last_link_id": 7,
6 "nodes": [
7 {
8 "id": 5,
9 "type": "CLIPLoader",
10 "pos": [
11 470,
12 190
13 ],
14 "size": [
15 520,
16 190
17 ],
18 "flags": [],
19 "order": 0,
20 "mode": 0,
21 "inputs": [],
22 "outputs": [
23 {
24 "name": "CLIP",
25 "type": "CLIP",
26 "links": [
27 5
28 ]
29 }
30 ],
31 "properties": {
32 "Node name for S&R": "CLIPLoader",
33 "cnr_id": "comfy-core",
34 "ver": "0.25.0",
35 "models": [
36 {
37 "name": "qwen3vl_4b_fp8_scaled.safetensors",
38 "url": "https://huggingface.co/Comfy-Org/Qwen3-VL/resolve/main/text_encoders/qwen3vl_4b_fp8_scaled.safetensors",
39 "directory": "text_encoders"
40 }
41 ]
42 },
43 "widgets_values": [
44 "qwen3vl_4b_fp8_scaled.safetensors",
45 "ideogram4",
46 "default"
47 ],
48 "widgets_values_named": {
49 "clip_name": "qwen3vl_4b_fp8_scaled.safetensors",
50 "type": "ideogram4",
51 "device": "default"
52 }
53 },
54 {
55 "id": 7,
56 "type": "PreviewAny",
57 "pos": [
58 1610,
59 190
60 ],
61 "size": [
62 590,
63 560
64 ],
65 "flags": [],
66 "order": 4,
67 "mode": 0,
68 "inputs": [
69 {
70 "name": "source",
71 "type": "*",
72 "link": 7
73 }
74 ],
75 "outputs": [
76 {
77 "name": "STRING",
78 "type": "STRING",
79 "links": null
80 }
81 ],
82 "properties": {
83 "Node name for S&R": "PreviewAny",
84 "cnr_id": "comfy-core",
85 "ver": "0.25.0"
86 },
87 "widgets_values": [],
88 "widgets_values_named": []
89 },
90 {
91 "id": 8,
92 "type": "LoadImage",
93 "pos": [
94 470,
95 450
96 ],
97 "size": [
98 520,
99 550
100 ],
101 "flags": [],
102 "order": 1,
103 "mode": 0,
104 "inputs": [],
105 "outputs": [
106 {
107 "name": "IMAGE",
108 "type": "IMAGE",
109 "links": [
110 6
111 ]
112 },
113 {
114 "name": "MASK",
115 "type": "MASK",
116 "links": null
117 }
118 ],
119 "properties": {
120 "Node name for S&R": "LoadImage",
121 "cnr_id": "comfy-core",
122 "ver": "0.25.0"
123 },
124 "widgets_values": [
125 "example.png",
126 "image"
127 ],
128 "widgets_values_named": {
129 "image": "example.png",
130 "upload": "image"
131 }
132 },
133 {
134 "id": 9,
135 "type": "MarkdownNote",
136 "pos": [
137 -150,
138 190
139 ],
140 "size": [
141 570,
142 820
143 ],
144 "flags": [],
145 "order": 2,
146 "mode": 0,
147 "inputs": [],
148 "outputs": [],
149 "properties": [],
150 "widgets_values": [
151 "**Input Assets**\n- [example.png](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/example.png) — required workflow input (`input/example.png`)\n\n- [#14298](https://github.com/Comfy-Org/ComfyUI/pull/14298)\n\n## Model Links\n\n- [Hugging Face:Comfy-Org/Qwen3-VL](https://huggingface.co/Comfy-Org/Qwen3-VL)\n\n**ModelScope mirrors** (alternative repo location)\n\n- [ModelScope:Comfy-Org/Qwen3-VL](https://modelscope.cn/models/Comfy-Org/Qwen3-VL)\n\n**text_encoders**\n\n- [qwen3vl_4b_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/Qwen3-VL/resolve/main/text_encoders/qwen3vl_4b_fp8_scaled.safetensors) (4.88 GB)\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ └── 📂 text_encoders/\n│ └── qwen3vl_4b_fp8_scaled.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"
152 ],
153 "widgets_values_named": {
154 "text": "**Input Assets**\n- [example.png](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/example.png) — required workflow input (`input/example.png`)\n\n- [#14298](https://github.com/Comfy-Org/ComfyUI/pull/14298)\n\n## Model Links\n\n- [Hugging Face:Comfy-Org/Qwen3-VL](https://huggingface.co/Comfy-Org/Qwen3-VL)\n\n**ModelScope mirrors** (alternative repo location)\n\n- [ModelScope:Comfy-Org/Qwen3-VL](https://modelscope.cn/models/Comfy-Org/Qwen3-VL)\n\n**text_encoders**\n\n- [qwen3vl_4b_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/Qwen3-VL/resolve/main/text_encoders/qwen3vl_4b_fp8_scaled.safetensors) (4.88 GB)\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ └── 📂 text_encoders/\n│ └── qwen3vl_4b_fp8_scaled.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"
155 },
156 "color": "#222",
157 "bgcolor": "#000"
158 },
159 {
160 "id": 10,
161 "type": "TextGenerateV2",
162 "pos": [
163 1100,
164 200
165 ],
166 "size": [
167 400,
168 476
169 ],
170 "flags": [],
171 "order": 3,
172 "mode": 0,
173 "inputs": [
174 {
175 "name": "clip",
176 "type": "CLIP",
177 "link": 5
178 },
179 {
180 "name": "images.image1",
181 "shape": 7,
182 "type": "IMAGE",
183 "link": 6
184 },
185 {
186 "name": "images.image2",
187 "shape": 7,
188 "type": "IMAGE",
189 "link": null
190 },
191 {
192 "name": "videos.video1",
193 "shape": 7,
194 "type": "IMAGE",
195 "link": null
196 },
197 {
198 "name": "audios.audio1",
199 "shape": 7,
200 "type": "AUDIO",
201 "link": null
202 },
203 {
204 "name": "system_prompt",
205 "shape": 7,
206 "type": "STRING",
207 "link": null
208 }
209 ],
210 "outputs": [
211 {
212 "name": "generated_text",
213 "type": "STRING",
214 "links": [
215 7
216 ]
217 },
218 {
219 "name": "thinking",
220 "type": "STRING",
221 "links": null
222 }
223 ],
224 "properties": {
225 "Node name for S&R": "TextGenerateV2"
226 },
227 "widgets_values": [
228 "what's in the images",
229 512,
230 "on",
231 0.7,
232 64,
233 0.95,
234 0.05,
235 1.05,
236 0,
237 0,
238 "randomize",
239 false,
240 true,
241 "auto"
242 ],
243 "widgets_values_named": {
244 "prompt": "what's in the images",
245 "max_length": 512,
246 "sampling_mode": "on",
247 "sampling_mode.temperature": 0.7,
248 "sampling_mode.top_k": 64,
249 "sampling_mode.top_p": 0.95,
250 "sampling_mode.min_p": 0.05,
251 "sampling_mode.repetition_penalty": 1.05,
252 "sampling_mode.presence_penalty": 0,
253 "sampling_mode.seed": 0,
254 "sampling_mode.seed.control_after_generate": "randomize",
255 "thinking": false,
256 "use_default_template": true,
257 "mtp": "auto"
258 }
259 }
260 ],
261 "links": [
262 [
263 5,
264 5,
265 0,
266 10,
267 0,
268 "CLIP"
269 ],
270 [
271 6,
272 8,
273 0,
274 10,
275 1,
276 "IMAGE"
277 ],
278 [
279 7,
280 10,
281 0,
282 7,
283 0,
284 "STRING"
285 ]
286 ],
287 "groups": [],
288 "config": [],
289 "extra": {
290 "ds": {
291 "scale": 0.39658169337787746,
292 "offset": [
293 1092.662154695437,
294 716.9945996853813
295 ]
296 },
297 "frontendVersion": "1.57.6",
298 "VHS_latentpreview": false,
299 "VHS_latentpreviewrate": 0,
300 "VHS_MetadataImage": true,
301 "VHS_KeepIntermediate": true
302 },
303 "version": 0.4
304}