A comprehensive AI Writing Intelligence Suite built with Next.js 14, TailwindCSS, Shadcn UI, and HuggingFace Inference API.
- Quick access to all tools
- Clean, modern interface
- Summary - AI-generated summaries using BART
- Sentiment - Positive/Negative classification
- Keywords - Automatic extraction
- Emotions - Joy, Anger, Sadness, Fear breakdown
- Toxicity - Content safety score
- Tone & Style - Writing classification
- Flesch Reading Ease
- Gunning Fog Index
- SMOG Index
- Grade Level Estimate
- SEO Score (0-100)
- Keyword Density
- Passive Voice %
- Sentence Length Analysis
- Improvement Recommendations
- 📧 Professional Email
- 🐦 Tweet
- 💼 Formal Business
- ✨ Creative
- 👶 ELI5 (Explain Like I'm 5)
- 📏 Shorten
- 📖 Expand
- LocalStorage-based storage
- View past analyses
- Delete entries
- Framework: Next.js 14 (App Router)
- Language: TypeScript
- Styling: TailwindCSS
- UI: Shadcn UI
- AI: HuggingFace Inference API
Important: This project requires
@huggingface/inferencev4.13.5 or later. Earlier versions (v2.x) use the deprecatedapi-inference.huggingface.coendpoint which no longer works. The current version automatically uses the newrouter.huggingface.coendpoint.
- Node.js 18.17+
- HuggingFace API key
npm installCreate .env.local:
HF_API_KEY=your_huggingface_api_keynpm run devnpm run build
npm start- Push to GitHub
- Import project in Vercel
- Add
HF_API_KEYenvironment variable - Deploy
| Feature | Model |
|---|---|
| Summarization | facebook/bart-large-cnn |
| Sentiment | distilbert-base-uncased-finetuned-sst-2-english |
| Keywords | ml6team/keyphrase-extraction-kbir-inspec |
| Emotions | j-hartmann/emotion-english-distilroberta-base |
| Toxicity | unitary/toxic-bert |
| Text Generation | Qwen/Qwen2.5-Coder-32B-Instruct |
MIT