Prompt Templates in awesome-gpt-image-2: 21 Industrial-Grade Visual Generation Categories

awesome-gpt-image-2 provides over 380 production-ready prompts organized into 21 high-level template families, each offering both plain-text patterns for manual use and structured JSON schemas for programmatic API calls.

The freestylefly/awesome-gpt-image-2 repository delivers a comprehensive prompt engineering library designed for consistent, reproducible image generation across diverse visual domains. Each template targets specific industry use-cases—from mobile UI mockups to historical artwork—ensuring optimal output quality through structured input constraints defined in docs/templates.md.

Dual-Format Template Architecture

Every prompt template in awesome-gpt-image-2 follows a dual-format design to accommodate both human operators and automated agents. The text format provides natural-language patterns with bracketed placeholders (e.g., [product type], [visual style]) for direct copy-paste workflows. The JSON-advanced format offers structured schemas optimized for API integration, featuring typed fields like "type", "product", and "constraints" that map directly to the generation backend.

This architecture is centrally documented in docs/templates.md, which serves as the master reference for all 13 primary template families. The repository routes these prompts through src/apimartClient.js to the generation endpoint at api/generate-image.js, which accepts both string and JSON payloads.

Interface and Commercial Design Templates

UI & Interface

The UI & Interface template generates high-fidelity mobile and web application mockups. The text pattern follows: 为[产品类型]生成一张[平台]界面图。核心功能：… 视觉风格：… 布局：… 输出：高保真UI截图…. The JSON equivalent specifies "type": "UI Screenshot" with nested platform and product fields. This template ensures consistent aspect ratios, readable typography, and platform-specific design patterns for iOS, Android, and web applications.

Product & E-Commerce

Designed for online retail, the Product & E-Commerce template produces hero images and shop-front shots. The structure requires specifying the product name, selling points, scene context, and camera angles. The JSON variant includes a product object with material, angle, and finish properties, plus a copywriting node for overlay text like pricing badges and slogans.

Brand & Identity

The Brand & Identity template constructs complete visual systems including logos, auxiliary graphics, and color palettes. Prompts specify brand keywords and deliverables such as "Logo direction" and "application demonstrations." This template enforces consistency across all generated brand assets through constrained style parameters.

Poster & Layout

For marketing materials, the Poster & Layout template handles campaign ads, movie posters, and typography-driven compositions. The pattern requires defining the subject (event, product, or film), main visual elements, layout structure, and color scheme. Output specifications mandate high-resolution formats suitable for social media distribution.

Infographic & Data-Viz

The Infographic & Data-Viz template creates structured information graphics, timelines, and charts. It enforces hierarchical information architecture through mandatory sections: header zones, data visualization areas, and summary blocks. The template supports various chart types and requires explicit definition of target audiences to ensure appropriate complexity levels.

Artistic and Narrative Content Templates

Photography & Realism

The Photography & Realism template produces hyper-realistic images mimicking professional camera work. It accepts detailed photographic parameters including lens specifications (35mm, 85mm), depth of field settings, and lighting conditions. The text pattern structures inputs as shooting subjects, locations, and technical camera settings to generate street photography, portraits, and product shots with photorealistic textures.

Illustration & Art

For stylized artwork, the Illustration & Art template supports concept art, cover illustrations, and editorial drawings. Parameters include art style references (e.g., "Studio Ghibli inspired"), line quality specifications, and color palette constraints. This template maintains consistency in stroke weight and compositional balance across generated pieces.

Character & Figure

The Character & Figure template generates character concept sheets, action-reference grids, and 3D toy models. It requires detailed physical descriptions including age, hairstyle, and clothing, alongside personality keywords. The template supports multi-panel outputs (such as 4×4 action grids) with consistency constraints ensuring identical proportions and facial features across all panels.

Scene & Narrative

Scene & Narrative templates construct plot-driven compositions with explicit temporal and spatial context. The pattern requires defining the story theme, time period, location, main event, protagonist details, and atmospheric mood. This ensures generated images convey specific narrative tension and emotional tone suitable for storyboarding and concept development.

History & Traditional

The History & Traditional template produces historically accurate artwork reflecting specific dynasties or cultural periods (such as Tang Dynasty settings). It constrains costumes, architecture, and social contexts to maintain period accuracy. Parameters include character identity, clothing details, and scene types (palace vs. marketplace) to ensure cultural and historical authenticity.

Spatial and Document Visualization

Architecture & Space

Architecture & Space templates render realistic interior and exterior architectural visualizations. Inputs specify space type (residential, commercial, landscape), functional purpose, material palettes, and lighting conditions. The template outputs photorealistic renderings suitable for client presentations and design reviews.

Document & Publishing

For editorial workflows, the Document & Publishing template layouts magazines, menus, and report spreads. It controls structural elements including column counts, margins, and content modules (cover zones, body text areas). The template prioritizes readability and typographic hierarchy for professional publishing standards.

Other Applications

The Other Applications template serves as a flexible catch-all for specialized use-cases including concept product boards, A/B testing scenarios, and custom generation tasks. It provides a structured framework for defining unique objectives while maintaining the repository's anti-pitfall guidance and output constraints.

Production Implementation

To utilize these templates programmatically, the repository provides src/apimartClient.js to handle prompt transmission and api/generate-image.js to process requests. The endpoint accepts both text strings and JSON objects, routing them to the generation backend.

Text Prompt Example: UI Generation

为移动金融 App 生成一张 iOS 界面图。
核心功能：账户概览、实时交易、AI 投资建议。
视觉风格：极简，主色蓝色，强调色橙色。
布局：顶部导航 + 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。

JSON Prompt Example: E-Commerce Product

{
  "type": "E-commerce Hero Image",
  "product": {
    "name": "Noise Cancelling Headphones",
    "material": "Matte black finish with metallic accents",
    "angle": "3/4 profile, floating slightly"
  },
  "setting": {
    "background": "Minimalist studio setup, soft gray gradient",
    "lighting": "Softbox overhead, sharp rim light on edges"
  },
  "copywriting": {
    "badges": ["NEW", "$299"],
    "slogan": "Silence the World"
  },
  "constraints": "Commercial photography quality, hyper‑realistic textures"
}

Text Prompt Example: Character Action Grid

设计游戏角色动作分解参考表。
风格：黑白线稿，背景纯净。
版式：4×4 网格，共 16 个等尺寸面板。
角色一致性：所有面板使用同一角色，保持脸型、服装、比例一致。
每格结构：
- 顶部：动作标题
- 中央：完整身体动作姿态
- 底部：3‑4 行动作说明
- 叠加：方向箭头、旋转箭头或运动轨迹线
约束：不要复杂背景，不要新增角色，不要彩色干扰。
输出：清晰可读、可用于动画/游戏动作参考的角色动作表。

Style Library Integration

The repository includes data/style-library.json, a curated reference database that can be invoked within any template via the style object (e.g., style.theme). This centralized style dictionary ensures consistency across multiple generation requests and provides pre-validated aesthetic parameters that complement the structural constraints defined in the prompt templates.

Summary

• 21 high-level template collections cover UI design, data visualization, photography, character art, architectural rendering, and publishing layouts.
• Dual-format architecture provides human-readable text patterns and structured JSON schemas for every template.
• Master documentation resides in docs/templates.md, defining specific parameter patterns and output constraints for each family.
• API integration routes through src/apimartClient.js and api/generate-image.js, supporting both text and JSON payloads.
• Style consistency is maintained via data/style-library.json for cross-template aesthetic alignment.

Frequently Asked Questions

How do I choose between text and JSON prompt formats in awesome-gpt-image-2?

Use the text format for manual, one-off generations where you can naturally describe visual requirements in sentences. The JSON format is optimal for automated workflows, agent-based systems, or when you need precise programmatic control over parameters like camera angles, color codes, or layout grids. According to the source code in api/generate-image.js, both formats parse through identical validation pipelines but JSON enables stricter type checking.

What file contains the complete list of prompt templates?

The complete catalog of prompt templates is defined in docs/templates.md at the repository root. This file contains the source definitions for all 13 primary template families, including specific anchor links (e.g., #tpl-ui, #tpl-character) that direct you to implementation examples for each category. The file specifies both the Chinese text patterns and their JSON equivalents.

Can I combine multiple template types in a single generation request?

Each template is designed as a discrete generation unit targeting specific visual outcomes (e.g., UI screenshots vs. character sheets). While you cannot directly merge template schemas in a single API call, the Other Applications template provides a flexible framework for hybrid requests. For complex multi-asset projects, chain multiple calls through src/apimartClient.js using the JSON format to maintain consistency across outputs.

How does the style library integrate with prompt templates?

The data/style-library.json file provides a centralized reference for aesthetic parameters that can be referenced inside any template's JSON structure using the style field (e.g., "style": { "theme": "minimalist", "palette": "monochrome" }). This ensures that visual elements like color schemes and lighting conditions remain consistent across different template families when generating related assets for a single project.

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