13 Categories of Prompt Templates in awesome-gpt-image-2: A Complete Guide

The freestylefly/awesome-gpt-image-2 repository organizes its image generation library into 13 distinct prompt template categories—including UI, Infographic, Poster, Product, Brand, Architecture, Photo, Illustration, Character, Scene, History, Document, and Other—each providing structured constraints for specific visual outputs.

The awesome-gpt-image-2 project delivers a comprehensive taxonomy for AI image generation through meticulously designed prompt templates. According to the source definitions in docs/templates.md, these 13 categories enable developers to generate everything from high-fidelity user interfaces to historical reconstructions with precise stylistic control.

The 13 Template Categories Overview

The template system divides visual generation tasks into 13 specialized domains. Each category includes both plain-text skeletons and JSON-structured variants for programmatic consumption.

• UI (用户界面) – Generates high-fidelity interface mockups for iOS, Android, and web applications with defined color schemes and layout constraints.
• Infographic (信息可视化) – Creates data-driven visualizations including flowcharts, comparison tables, timelines, and scientific scales.
• Poster (海报/排版) – Designs promotional materials with dominant typography and central visual elements for editorial layouts.
• Product – Renders isolated product photography with specific lighting, materials, and background controls.
• Brand – Produces brand identity assets including logos, business cards, and style guide elements.
• Architecture – Generates building exteriors, interior spaces, and urban planning visualizations with structural accuracy.
• Photo – Creates photorealistic imagery with specified camera angles, lighting conditions, and lens effects.
• Illustration – Produces artistic drawings in various styles including vector art, watercolor, and technical diagrams.
• Character – Designs human figures, avatars, and mascots with defined poses, expressions, and costume details.
• Scene – Composes environmental backgrounds and settings with atmospheric perspective and lighting.
• History – Reconstructs historical events, ancient architecture, and period-specific visual documentation.
• Document – Generates formal documents, certificates, resumes, and paperwork with proper typography.
• Other – Catch-all category for specialized templates not covered by the primary classifications.

Template Structures for Key Categories

Three categories form the core usage patterns for most developers: UI, Infographic, and Poster. Each provides structured plaintext and JSON formats as defined in docs/templates.md.

UI (User Interface) Templates

UI templates generate platform-specific interface screenshots with strict fidelity requirements.

Plain-text template structure:

# UI template (plain‑text)

为[产品类型]生成一张[平台，如 iOS/Android/Web]界面图。
核心功能：[功能点A]、[功能点B]、[功能点C]。
视觉风格：[极简/科技/拟物]，主色[颜色]，强调色[颜色]。
布局：[顶部导航/双栏/卡片流]，信息层级清晰，留白充足。
输出：高保真UI截图，文字清晰可读，比例[9:16/16:9]。

JSON programmatic variant:

{
  "type": "UI Screenshot",
  "platform": "iOS",
  "product": "Fitness App",
  "layout": "Card‑based feed with bottom tab bar",
  "style": {
    "theme": "Dark Mode",
    "primary_color": "Neon Green",
    "typography": "Clean sans‑serif"
  },
  "content": {},
  "constraints": "High fidelity, readable text, 9:16 aspect ratio"
}

Infographic Templates

Infographic templates enforce structured information hierarchy for data visualization outputs.

Plain-text template:

# Infographic template (plain‑text)

生成[主题]信息图，目标读者为[人群]。
结构：标题区 + [3‑5]个模块（每模块含图标、短标题、1‑2句说明 …）。
图表类型：[流程图/对比图/关系图/时间线]。
风格：[专业报告/科普插画/儿童教育]，主色[颜色]，背景[浅色/深色]。
输出：信息层级清晰、可读性高的中文信息图。

JSON specification:

{
  "type": "Infographic",
  "topic": "Urban Metabolism",
  "audience": "General Public",
  "structure": {},
  "style": {},
  "constraints": "No cyberpunk, no gibberish text, strict structural layout"
}

Poster Templates

Poster templates prioritize compositional balance and typographic dominance for marketing materials.

Plain-text format:

# Poster template (plain‑text)

设计一张[活动/产品/电影]海报，主题为[主题词]。
主视觉：[主体元素]，标题文案：[标题]，副标题：[副标题]。
版式：[居中/左对齐/对角构图]，风格：[复古/未来/极简]。
色彩：[主色 + 辅色]，氛围：[情绪关键词]。
输出：可用于社媒传播的高分辨率海报。

JSON structure:

{
  "type": "Poster",
  "theme": "Future City",
  "visuals": {
    "subject": "Silhouette of a sleek skyscraper skyline",
    "style": "Minimalist black‑white with neon accent",
    "color_palette": "Black, White, Neon Blue"
  },
  "typography": {},
  "constraints": "High‑resolution, 9:16 aspect ratio, title must be dominant"
}

Implementation in the Source Code

The 13-category system is implemented across three critical files in the repository.

Template Definitions: docs/templates.md contains the complete specifications for all 13 categories, including parameter schemas and example outputs for each template type.

Frontend Interface: src/main.jsx renders the category selection UI, allowing users to browse the 13 template types and preview generation results before submitting requests.

Backend Processing: api/generate-image.js receives the filled template payloads—whether plain-text or JSON—and routes them to the image generation model with appropriate parsing logic for each of the 13 category types.

Summary

• The awesome-gpt-image-2 repository defines 13 distinct prompt template categories spanning UI, Infographic, Poster, Product, Brand, Architecture, Photo, Illustration, Character, Scene, History, Document, and Other.
• Each category provides dual-format templates (plain-text and JSON) stored in docs/templates.md for both human readability and programmatic integration.
• UI templates enforce platform-specific constraints for mobile and web interfaces, while Infographic templates prioritize data hierarchy and Poster templates focus on typographic dominance.
• The implementation splits responsibilities between src/main.jsx (frontend selection), api/generate-image.js (backend processing), and docs/templates.md (source definitions).

Frequently Asked Questions

What are all 13 categories available in awesome-gpt-image-2?

The 13 categories are UI (User Interface), Infographic, Poster, Product, Brand, Architecture, Photo, Illustration, Character, Scene, History, Document, and Other. These cover everything from mobile app mockups to historical reconstructions, with the "Other" category handling specialized edge cases not fitting the primary taxonomies.

How do the JSON templates differ from plain-text versions?

JSON templates provide structured key-value pairs for programmatic consumption, enabling automated form filling and validation through api/generate-image.js, while plain-text versions use bracketed placeholders like [产品类型] for manual editing and quick prototyping in conversational interfaces.

Which source file contains the complete category definitions?

All 13 category specifications, including parameter schemas and example prompts, are documented in docs/templates.md at the repository root. This file serves as the single source of truth for both the frontend selector in src/main.jsx and the backend generation logic.

Can I combine multiple categories in a single prompt?

While each template in docs/templates.md is designed for a specific category, the "Other" classification accommodates hybrid requests. For strict category adherence, the backend endpoint api/generate-image.js processes one category type per request to ensure constraint enforcement specific to that template family.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}