WriteLike

WriteLike helps writers, editors, and researchers quickly understand the tone, style, and structure of any text. Powered by spaCy and AI, it identifies key linguistic patterns, sentiment, and readability metrics, turning complex text into actionable insights — making writing analysis faster, easier, and more effective.

Enter text in English. (Simplified Chinese is supported experimentally.)

Analysis:

  1. Overall Style & Tone:
    The frequent lemmas like "generalize," "war," "true," and "death" suggest a serious, reflective, analytical, or even philosophical tone. It leans formal/academic rather than conversational.

  2. Writing Techniques & Effects:

Pacing: High oscillation ratios for both sentence and token length suggest varied pacing. Concise statements mix with longer, complex sentences to create rhythm and emphasis.

Clarity: With ~1.75 clauses per sentence, the writing stays digestible while retaining nuance.

Guidance: To emulate this style:

  • Vary sentence lengths for dynamic pacing.
  • Balance concise and complex vocabulary.
  • Explore abstract themes (truth, conflict, mortality).
  • Maintain a serious and analytical tone.

While you wait -

What WriteLike does

WriteLike looks at the “shape” of your writing — things like sentence length, rhythm, tone, and word choice — and turns those patterns into clear, human-friendly feedback. It’s not about saying if writing is “good” or “bad,” but showing how it works: punchy and direct, reflective and flowing, descriptive, or action-driven.


What it’s for

The goal is to help you ask: “Why does this passage feel the way it does?”WriteLike highlights the components of style so you can borrow rhythms, structures, and tones into your own work — while keeping your own voice. It’s a tool for exploration, learning, and writing more intentionally.


Why it works

Behind the scenes, WriteLike blends linguistic analysis (spaCy, TextBlob, statistical features) with AI interpretation (Google Gemini). Think of it as a translator between data and storytelling: the numbers describe what’s happening, and the AI explains what it means for your style.


Technical at a glance

  • Language Detection: langdetect (English + Simplified Chinese, experimental)
  • NLP: spaCy 3.8.7 (en_core_web_sm, zh_core_web_sm), TextBlob for sentiment
  • Features: sentence & token lengths, clause density, POS & dependency distributions, verb tense, lemma motifs, morphology samples
  • Generative AI: Google Gemini 2.5 API for style feedback
  • Backend: FastAPI microservice (Python) hosted on Render
  • Frontend: Next.js app fetching results via API

Learn more