工作流 /
图像 /
Nano Banana 2 Lite: 文生图
Nano Banana 2 Lite: 文生图
Nano Banana 2 Lite: Text to Image
使用超快Gemini 3.1 Flash-Lite图像模型，根据文本描述生成图像，无需文件输入，专为快速创建和迭代设计。此工作流根据您的文本提示词生成一个或多个已生成图像，支持交错文本和图像输入。非常适合快速概念可视化、快速原型设计和迭代设计探索。
图像
2 个节点 16 次使用 ComfyUI ≥ 0.27.0 ComfyUI 3 KB
节点构成
GeminiNanoBanana2V2
× 1
SaveImageAdvanced
× 1
官方标注的模型
Nano Banana 2
工作流 JSON
和上面「下载工作流 JSON」是同一份文件。
1{
2 "id": "23df7256-245b-422d-8f14-d5cb6ee5f619",
3 "revision": 0,
4 "last_node_id": 7,
5 "last_link_id": 3,
6 "nodes": [
7 {
8 "id": 3,
9 "type": "GeminiNanoBanana2V2",
10 "pos": [
11 -280,
12 240
13 ],
14 "size": [
15 400,
16 440
17 ],
18 "flags": [],
19 "order": 0,
20 "mode": 0,
21 "inputs": [
22 {
23 "label": "image_1",
24 "name": "model.images.image_1",
25 "shape": 7,
26 "type": "IMAGE",
27 "link": null
28 },
29 {
30 "label": "files",
31 "name": "model.files",
32 "shape": 7,
33 "type": "GEMINI_INPUT_FILES",
34 "link": null
35 }
36 ],
37 "outputs": [
38 {
39 "name": "IMAGE",
40 "type": "IMAGE",
41 "links": [
42 2
43 ]
44 },
45 {
46 "name": "STRING",
47 "type": "STRING",
48 "links": null
49 },
50 {
51 "name": "thought_image",
52 "type": "IMAGE",
53 "links": null
54 }
55 ],
56 "properties": {
57 "Node name for S&R": "GeminiNanoBanana2V2"
58 },
59 "widgets_values": [
60 "90s vintage F1 editorial poster, extreme low-angle close-up. A sleek electric blue and silver F1 car tearing through the frame, driver in an iconic yellow-black helmet. Hyperspeed motion blur, aggressive speed streaks, burning tire smoke. Massive bold yellow typography \"Nano Banana 2 Lite\" dominating a pale blue sky. Avant-garde graphic layout, halftone print texture, folded paper creases, high-contrast 35mm film grain. Minimalist micro-text and barcode at the bottom. Dynamic, ultra-fast, cinematic retro design, no logos",
61 "Nano Banana 2 Lite",
62 "1:1",
63 "1K",
64 "MINIMAL",
65 129222452393446,
66 "randomize",
67 "IMAGE",
68 "You are an expert image-generation engine. You must ALWAYS produce an image.\nInterpret all user input—regardless of format, intent, or abstraction—as literal visual directives for image composition.\nIf a prompt is conversational or lacks specific visual details, you must creatively invent a concrete visual scenario that depicts the concept.\nPrioritize generating the visual representation above any text, formatting, or conversational requests.",
69 1,
70 0.95
71 ],
72 "color": "#432",
73 "bgcolor": "#653"
74 },
75 {
76 "id": 4,
77 "type": "SaveImageAdvanced",
78 "pos": [
79 150,
80 240
81 ],
82 "size": [
83 570,
84 450
85 ],
86 "flags": [],
87 "order": 1,
88 "mode": 0,
89 "inputs": [
90 {
91 "name": "images",
92 "type": "IMAGE",
93 "link": 2
94 }
95 ],
96 "outputs": [
97 {
98 "name": "images",
99 "type": "IMAGE",
100 "links": null
101 }
102 ],
103 "properties": [],
104 "widgets_values": [
105 "Nano_banana_2_lite",
106 "png",
107 "8-bit",
108 "sRGB"
109 ]
110 }
111 ],
112 "links": [
113 [
114 2,
115 3,
116 0,
117 4,
118 0,
119 "IMAGE"
120 ]
121 ],
122 "groups": [],
123 "config": [],
124 "extra": {
125 "ds": {
126 "scale": 0.42409761837248483,
127 "offset": [
128 1069.0614446568331,
129 454.45475584116093
130 ]
131 },
132 "frontendVersion": "1.47.6"
133 },
134 "version": 0.4
135}