An AI-powered desktop app for one-click pet content creation — from article generation and image sourcing to Word document export, all in one place.
- 🤖 AI-Powered Writing — Leverages the DeepSeek large language model to generate Toutiao-style pet articles based on topic, pet type, and writing style
- 🖼️ Auto Image Sourcing — Integrated with the free Pexels image library; AI automatically extracts keywords to search and download high-quality images
- 📄 Word Export — Export articles with images as a formatted
.docxWord document in one click - 🎛️ Flexible Configuration — Customizable DeepSeek model, word count, image count, writing style, and more
- 🖥️ Native Desktop Experience — Built with Wails3 for a native window, low memory footprint, and no browser required
| Style | Description |
|---|---|
| 📚 Science | Professional pet knowledge articles |
| 📋 Guide | Practical pet care tips |
| 📖 Story | Heartwarming pet stories |
| ⭐ Review | Pet product reviews & recommendations |
| 📰 News | Pet industry news & trends |
🐕 Dog · 🐈 Cat · 🐰 Rabbit · 🐹 Hamster · 🐦 Bird · 🐟 Fish · 🦎 Reptile · 🐾 Other
User inputs writing requirements
↓
AI generates article (DeepSeek)
↓
AI extracts image search keywords
↓
Search images (Pexels API)
↓
Download images → Embed in document
↓
Generate Word document (.docx)
↓
Done! Article + Images + Word Export
| Layer | Technology |
|---|---|
| Desktop | Wails3 (Go + WebView2) |
| Backend | Go 1.25 |
| Frontend | React 18 + TypeScript + Vite |
| AI Engine | DeepSeek Chat Model (via CloudWeGo Eino) |
| Image API | Pexels API |
| Doc Generation | Native OOXML builder (pure Go, zero dependencies) |
pet-content-creator/
├── main.go # Application entry point
├── go.mod / go.sum # Go dependencies
├── Taskfile.yml # Build tasks
├── backend/
│ ├── config.go # Configuration management (API keys, etc.)
│ ├── agent.go # AI article generation service
│ ├── pexels.go # Pexels image search/download
│ ├── word.go # Word document generation
│ └── creator.go # Orchestration service (unified entry)
├── frontend/
│ ├── index.html # Entry page
│ ├── package.json # Frontend dependencies
│ ├── vite.config.ts # Vite configuration
│ └── src/
│ ├── main.tsx # React entry
│ └── App.tsx # Main UI (form + results)
└── build/ # Build artifacts
- Go 1.24+
- Node.js 18+
- Wails3 CLI (see installation below)
- DeepSeek API Key — get one at platform.deepseek.com
- Pexels API Key (free) — register at pexels.com/api
go install github.com/wailsapp/wails/v3/cmd/wails3@latest# Clone the repo
git clone <your-repo-url> pet-content-creator
cd pet-content-creator
# Install frontend dependencies
cd frontend
npm install
cd ..
# Development mode (with hot-reload)
wails3 dev
# Production build
wails3 buildBuild artifacts will be placed in the build/ directory.
On first launch, click the ⚙️ Config button in the top-right corner and fill in:
- DeepSeek API Key: format
sk-xxx - DeepSeek Model: default
deepseek-chat, can also usedeepseek-reasoner - Pexels API Key: free after registration
Config is saved at ~/.pet-content-creator/config.json.
- Launch the app and enter a writing topic in the left panel (e.g., "Summer Heat Safety Guide for Dogs")
- Select a pet type and writing style
- Set the target word count (200–5000) and number of images (0–10)
- Optionally add keyword hints to guide the AI
- Click 🚀 Start Creating and wait 10–30 seconds
- The right panel shows the article preview and images; click 📄 View Word Document to open the exported
.docxfile
Generated Word files are saved under Documents\PetContentCreator\.
Handles reading and writing of API key and model configuration. API keys are masked when returned to the frontend.
Wraps DeepSeek ChatModel calls, builds a Toutiao-style system prompt, and parses the AI response into title, body, and image search keywords.
Calls the Pexels API to search for high-quality royalty-free images. Supports keyword search and image download (10 MB limit).
Pure Go implementation of OOXML .docx file generation (ZIP-based format). Built-in styles with Microsoft YaHei font, heading layout, and body text styling.
Orchestration service that chains the full pipeline: AI generation → image search → image download → Word generation.
The project also supports a headless server mode with HTTP API access:
# Build Docker image
task build:docker
# Run Docker container
task run:dockerMIT License
Star ⭐ and Issues 🐛 are welcome!