Perspective

API

NLP API for detecting toxic, obscene, insulting, or threatening text

Overview

Perspective API is a RESTful NLP service that analyzes text to return probability scores for harmful content categories like TOXICITY, OBSCENE, INSULT, and THREAT. It accepts JSON payloads with text input and returns structured JSON responses with 0-1 scale scores for each requested category. Key endpoints include the analyze endpoint for text evaluation. Use cases include moderating user-generated content (comments, forums), filtering harmful chat messages, enhancing educational platform safety, and automating content review to reduce manual effort.

Example Integration (JavaScript)

script.js JS


fetch('https://perspectiveapi.com')
  .then(res => res.json())
  .then(data => console.log(data))
  .catch(err => console.error(err));

Key Features

RESTful interface
JSON request/response
Multiple content safety categories
Probability score outputs
Multi-language support

Frequently Asked Questions

? Is Perspective API free to use?

Yes, it offers a free tier with limited monthly requests; paid plans are available for higher usage volumes.

? Does it require an API Key?

Yes, you need to obtain an API key from Google Cloud to authenticate all requests.

? What is the response format?

Responses are in JSON format, containing probability scores for each requested content category (e.g., TOXICITY, INSULT).

Top Alternatives

Tool Info

Pricing Freemium

Category Machine Learning

Platform Public API

Pros

⊕ Accurate Google NLP models
⊕ Easy integration
⊕ Detailed category scores
⊕ Scalable cloud-based service

Cons

⊖ Rate-limited free tier
⊖ Requires API key
⊖ Potential category bias
⊖ Cloud-only deployment