REST API v1.0 • Active

Integrate High-Precision
AI Detection API

Programmatically detect AI-generated text in real-time. Gain sentence-level probability scores, perplexity metrics, and multi-model classification with a simple HTTPS request.

Get Free API Key API Technical Specs

< 150ms Latency • 99.9% Uptime SLA

detect_request.js

Engineered for Accuracy & Scalability

Built to handle high-throughput workloads with sub-second response speeds.

Multi-Model Detection

Detects content generated by ChatGPT (GPT-4o, GPT-5), Claude Fable, Gemini 3.6, Llama 4, and custom fine-tuned LLMs with 99.4% precision.

Sentence-Level Scoring

Returns granular sentence-by-sentence probability scores, perplexity ratings, and burstiness indices so you know exactly which sections are AI-written.

Sub-150ms Edge Response

Served via globally distributed edge infrastructure with automatic load balancing to power real-time editors, LMS systems, and publisher CMS pipelines.

REST API Reference

Endpoint Specifications

Submit HTTP POST requests with raw text payload to receive instant JSON classification.

POST https://gpt-zero.com/api/detect

Required HTTP Headers

Authorization: Bearer YOUR_API_KEY

Content-Type: application/json

JSON Body Parameters

Field	Type	Description
text	string	Required. Raw text to analyze (100 to 100,000 characters).
model_version	string	Optional. Set to `latest` or `v1.0`.

JSON Response (HTTP 200 OK) ● 200 Success

{
  "status": "success",
  "data": {
    "ai_score": 0.994,
    "classification": "AI-Generated",
    "perplexity": 12.4,
    "burstiness": 3.8,
    "sentences": [
      {
        "id": 1,
        "text": "Artificial intelligence systems process huge amounts of data rapidly.",
        "score": 0.99,
        "class": "ai"
      },
      {
        "id": 2,
        "text": "I wrote this research paper on my computer yesterday.",
        "score": 0.02,
        "class": "human"
      }
    ]
  }
}

Bearer Authentication & TLS 1.3

Generate your API keys inside your developer dashboard. All API communications require TLS 1.3 encryption and Bearer authorization tokens. Never expose your secret key inside public client-side JavaScript.

Rate Limiting & Quota Rules

Standard accounts allow up to 60 requests per minute. Exceeding your plan's monthly limit returns HTTP status 429. Custom higher throughput & SLA quotas are available on Enterprise plans.

Developer FAQs

Which AI models are detected by the API?

Our detection model identifies outputs generated by ChatGPT (GPT-4o, GPT-5), Claude Fable, Gemini 3.6, Llama 4, and custom fine-tuned LLMs.

How do I obtain my API key?

Can we customize the AI detection score threshold?

Yes. Since our payload returns the exact probability float (0.0 to 1.0) and sentence classification, you can define custom policies.

Is there an official SDK for Node.js or Python?

You can make standard HTTP POST requests using any HTTP client (fetch, axios, requests, guzzle). Snippets are provided in our quick start guide.

Ready to Start Building?

Generate your API key in seconds and start detecting AI-generated content programmatically.

Create Free Account & Get Key

Integrate High-Precision AI Detection API