Back to Home View Source

Text Similarity API

Fuzzy matching + semantic analysis for flexible text comparison

A FastAPI service that detects typos using Levenshtein distance and understands meaning through transformer-based embeddings. Built for quiz apps, form validation, and flexible answer checking.

Try It Live

Tech Stack

Python
FastAPI
Sentence Transformers

all-MiniLM-L6-v2

RapidFuzz

How It Works

  1. 1.Normalize: Convert both inputs to lowercase
  2. 2.Fuzzy Check: Calculate Levenshtein distance ratio (catches typos)
  3. 3.Semantic Check: Encode both as embeddings, measure cosine similarity
  4. 4.Verdict: Pass if either threshold is met

Use Cases

  • •Quiz/trivia game answer validation
  • •Form input validation with typo tolerance
  • •Educational platforms for grading short answers
  • •Search systems with semantic understanding