# GPT Image Hub prompt

## 并排模型架构对比信息图
- URL: https://www.gptimagehub.com/zh-CN/library/side-by-side-model-architecture-comparison-infographic
- Generator URL: https://www.gptimagehub.com/zh-CN/generate?promptId=oss-ym-g2-15614
- Markdown URL: https://www.gptimagehub.com/api/agent/markdown?path=%2Fzh-CN%2Flibrary%2Fside-by-side-model-architecture-comparison-infographic
- Category: 信息图与教育
- Model: gpt-image-2
- Aspect ratio: 2:1
- Style: graphic-design
- Tags: 信息图, 技术图表, 架构, 对比图, 幻灯片, transformer
- Updated: 2026-10-27T13:35:53.098Z
- Created: 2026-10-27T13:35:53.098Z

生成用于技术展示的两个大模型架构并排对比图。可通过修改方括号内的模型名称、参数量和层维度进行定制。

### Full prompt
```text
{"type":"并排式AI模型架构对比信息图","style":"整洁的技术架构图、白底、细黑色轮廓线、圆角矩形、虚线标注框、颜色高亮、演示文稿幻灯片美学、矢量信息图","canvas":{"aspect_ratio":"2:1","resolution":"wide horizontal"},"title_row":{"left_title":"[DeepSeek V3/R1 (671 billion)]","right_title":"[DeepSeek V4 (1.2 trillion)]","left_title_color":"亮橙红色","right_title_color":"鲜蓝色"},"layout":{"columns":2,"sections":[{"title":"[DeepSeek V3/R1 (671 billion)]","position":"左半部","count":9,"labels":["Vocabulary size of 129k","FeedForward (SwiGLU) module","Intermediate hidden layer dimension of 2,048","MoE layer","Supported context length of 128k tokens","First 3 blocks use dense FFN with hidden size 18,432 instead of MoE","Sample input text","Embedding dimension of 7,168","128 heads"]},{"title":"[DeepSeek V4 (1.2 trillion)]","position":"右半部","count":9,"labels":["Vocabulary size of 160k","FeedForward (SwiGLU) module","Intermediate hidden layer dimension of 3,072","MoE layer","Supported context length of 256k tokens","First 3 blocks use dense FFN with hidden size 24,576 instead of MoE","Sample input text","Embedding dimension of 8,192","128 heads"]},{"title":"底部对比表格","position":"底部全宽","count":10,"labels":["Total parameters","Active parameters per token","Hidden size","Intermediate (FF)","Attention heads","Context length","Embedding dimension","Vocabulary size"]}]},"left_panel":{"background":"极浅灰色圆角矩形","main_stack":{"count":8,"blocks":["Tokenized text","Token embedding layer","RMSNorm 1","Multi-head Latent Attention","RMSNorm 2","MoE","Final RMSNorm","Linear output layer"]},"side_module":"左侧注意力模块附带RoPE","attention_block":{"label":"Multi-head Latent Attention","accent":"Latent字样使用橙红色"},"feedforward_inset":{"title":"FeedForward (SwiGLU) module","count":4,"blocks":["Linear layer","SiLU activation","Linear layer","Linear layer"],"diagram":"双分支相乘后投影"},"moe_inset":{"title":"MoE layer","count":5,"blocks":["top combine node","Feed forward","Feed forward","Router","expert count badge 256"],"details":"代表1个选中专家的黑色小方块、指向专家层的向上箭头、虚线分隔线"},"annotations":{"vocab":"Vocabulary size of 129k","ff_dim":"Intermediate hidden layer dimension of 2,048","context":"Supported context length of 128k tokens","dense_first_blocks":"First 3 blocks use dense FFN with hidden size 18,432 instead of MoE","resource_savings":"Resource savings: Model size is 671B but only 1 (shared) + 8 experts active per token; only 37B parameters are active per inference step"},"bottom_stats":{"count":8,"items":["Total parameters: 671B","Active parameters per token: 37B (1 + 8 experts)","Hidden size: 7,168","Intermediate (FF): 2,048","Attention heads: 128","Context length: 128k","Embedding dimension: 7,168","Vocabulary size: 129k"]}},"right_panel":{"background":"极浅蓝色圆角矩形","main_stack":{"count":8,"blocks":["Tokenized text","Token embedding layer","RMSNorm 1","Multi-head Latent Attention","RMSNorm 2","MoE","Final RMSNorm","Linear output layer"]},"side_module":"左侧注意力模块附带RoPE","attention_block":{"label":"Multi-head Latent Attention","accent":"Latent字样使用蓝色"},"feedforward_inset":{"title":"FeedForward (SwiGLU) module","count":4,"blocks":["Linear layer","SiLU activation","Linear layer","Linear layer"],"diagram":"与左侧结构相同"},"moe_inset":{"title":"MoE layer","count":5,"blocks":["top combine node","Feed forward","Feed forward","Router","expert count badge 384"],"details":"代表1个选中专家的黑色小方块、指向专家层的向上箭头、虚线分隔线、蓝色边框强调"},"annotations":{"vocab":"Vocabulary size of 160k","ff_dim":"Intermediate hidden layer dimension of 3,072","context":"Supported context length of 256k tokens","dense_first_blocks":"First 3 blocks use dense FFN with hidden size 24,576 instead of MoE","resource_savings":"Resource savings: Model size is 1.2T but only 1 (shared) + 8 experts active per token; only 52B parameters are active per inference step"},"bottom_stats":{"count":8,"items":["Total parameters: 1.2T","Active parameters per token: 52B (1 + 8 experts)","Hidden size: 8,192","Intermediate (FF): 3,072","Attention heads: 128","Context length: 256k","Embedding dimension: 8,192","Vocabulary size: 160k"]}},"global_notes":"生成高度对称的超详细Transformer模型架构对比图。左右各包含一个完整模型堆栈以及2个内嵌结构图（前馈模块与MoE专家层）。模块间附带连接箭头、细小技术文字与指引线。保持密集专业的幻灯片排版风格，左侧使用橙红色强调，右侧使用蓝色强调，底部附带横跨全宽的数据对比表。"}
```