Developer Docs

API Documentation

Production-Verified API

Complete API docs built from the real server contract, not placeholders.

This page documents the public API at /api/v1. It reflects the current API behavior, AI service integration, billing rules, and actual response formats so developers can copy examples and start building immediately.

Base URL

https://api.gpthumanizer.ai/api/v1

All public API requests go to this base path.

Authentication

X-API-Key

Get your API key from dashboard API panel.

Endpoints

5 public endpoints

credits, humanize, detect, writing/create, writing/refine.

Start Here

Generate an API key in your dashboard and store it on the server side.

Send JSON bodies and include X-API-Key in the request headers.

Parse the outer wrapper first: code, message, data, timestamp.

Handle 400, 401, 402, 429, and 500 errors explicitly in your client code.

Humanize

100 input words = 1 credit. Default model is balanced.

Detect

Costs 1 credit per detection request.

Writing Create

10 output tokens = 1 credit. Supports research mode.

Credits

Check remaining balance before expensive generation calls.

Request Basics

Required headers

X-API-Key: YOUR_API_KEY

Content-Type: application/json

General rules

All endpoints are synchronous from the client perspective. The service handles internal polling for you.
Use server-side API calls whenever possible. Do not expose secret API keys in public frontend code.
Humanize and detect reject inputs under 50 characters.
Writing endpoints bill on generated output tokens, not on input tokens.

Successful response wrapper

{
  "code": 200,
  "message": "Success",
  "data": {
    "request_id": "req_1711863600123_ab12cd34e",
    "humanized_text": "Rewritten text appears here"
  },
  "timestamp": 1711863600456
}

Error response wrapper

{
  "code": 402,
  "message": "Insufficient API credits. Required: 3 credits, Available: 0 credits.",
  "data": null,
  "timestamp": "2026-03-31T10:16:55.031Z",
  "path": "/api/v1/humanize",
  "method": "POST"
}

Contents

GET /credits POST /humanize POST /detect POST /writing/create POST /writing/refine Error handling

Balance

Get Remaining Credits

GET

/api/v1/credits

Use this endpoint to fetch the current remaining API credit balance for the API key. It is the safest way to check whether a user can afford a generation or detection request before you send it.

Authentication

Required

Requires X-API-Key.

Billing

Free

No credits are consumed by this endpoint.

Response note

unit = words

Current server response returns words as the unit even though the number is credit balance.

Response Fields

Field

Type

Required

Description

balance

number

Yes

Current remaining API credits balance.

unit

string

Yes

Current implementation returns words, even though balance represents API credits.

userId

string | number

Yes

Owner user ID for the API key.

Code Example

curl -X GET https://api.gpthumanizer.ai/api/v1/credits \
  -H "X-API-Key: YOUR_API_KEY"

Response Example

{
  "code": 200,
  "message": "Success",
  "data": {
    "balance": 96,
    "unit": "words",
    "userId": 1024
  },
  "timestamp": 1711863600456
}

Core Endpoint

Humanize Text

POST

/api/v1/humanize

Turn your text into natural writing. For a simpler estimate, you can think of it as 100 words = 1 credit based on 1 word ≈ 1.2 tokens. Actual billing still uses counted input tokens and rounds up.

Validation

>= 50 chars

Also capped at 40000 characters and 5000 input tokens.

Billing

100 words = 1 credit

Estimated using 1 word ≈ 1.2 tokens. Actual billing still uses counted input tokens, rounded up.

Default config

balanced + general + auto

All three are optional. Language auto-detects if omitted.

Supported models

free

Lowest-cost mode. Fast, but output quality is usually the weakest.

balanced

Recommended default. Best tradeoff between speed and quality.

enhanced

Highest-quality mode. Use when output quality matters most.

Supported styles

general

Default style for most content.

academic

Formal and academic wording.

casual

More conversational tone.

professional

Business-friendly and polished.

blog

Lighter, article-style writing.

Email-style phrasing and structure.

Recommended language codes

en-US

English

es-ES

Spanish

zh-CN

Chinese

Request Fields

Field

Type

Required

Description

text

string

Yes

Source text to humanize. At least 50 characters. Hard-capped at 40000 characters and 5000 input tokens.

model

string

Humanization model. Default is balanced.

style

string

Output writing style. Default is general.

language

string

Language code such as en-US or zh-CN. Default is auto (automatic language detection).

Code Example

curl -X POST https://api.gpthumanizer.ai/api/v1/humanize \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Artificial intelligence has revolutionized modern technology, but the wording still sounds too mechanical and repetitive.",
    "model": "balanced",
    "style": "general",
    "language": "en-US"
  }'

Response Fields

Field

Type

Required

Description

request_id

string

Yes

Public API request identifier.

upstream_request_id

string

Yes

Internal processing task identifier.

humanized_text

string

Yes

Final rewritten text.

usage.input_tokens

number

Yes

Input token count used for billing.

usage.output_tokens

number

Yes

Output token count for reference.

usage.credits_charged

number

Yes

Credits deducted for this call.

credits_remaining

number

Yes

Remaining API credits after billing.

ai_confidence

number

AI confidence score when available.

human_rate

number

Human-likeness score when available.

Response Example

{
  "code": 200,
  "message": "Success",
  "data": {
    "request_id": "req_1711863600123_ab12cd34e",
    "upstream_request_id": "4de55b32-fd3f-4c18-8c53-d00dcdd51a91",
    "humanized_text": "Artificial intelligence is reshaping modern technology at an extraordinary pace.",
    "usage": {
      "input_tokens": 268,
      "output_tokens": 241,
            "credits_charged": 3
    },
        "credits_remaining": 97,
    "ai_confidence": 0.18,
    "human_rate": 0.91
  },
  "timestamp": 1711863600456
}

Analysis

AI Detection

POST

/api/v1/detect

Detect whether a text looks AI-generated. The server validates text length, processes the detection request asynchronously, and returns AI and human probability scores. Each detection costs 1 credit.

Validation

>= 50 chars

Also capped at 40000 characters and 5000 input tokens.

Billing

1 credit

Each detection request costs exactly 1 credit.

Processing

Async

Detection runs asynchronously and is polled until completion.

Request Fields

Field

Type

Required

Description

text

string

Yes

Text to detect. At least 50 characters. Hard-capped at 40000 characters and 5000 input tokens.

Code Example

curl -X POST https://api.gpthumanizer.ai/api/v1/detect \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Your text to analyze for AI detection. Make sure it is long enough to pass minimum-length validation."
  }'

Response Fields

Field

Type

Required

Description

request_id

string

Yes

Public API request identifier.

is_ai_generated

boolean

Yes

Whether the text is predicted to be AI-generated.

ai_probability

number

Yes

Probability score for AI-generated content (0-100).

human_probability

number

Yes

Probability score for human-written content (0-100).

confidence

number

Yes

Confidence score for the prediction (0-1).

usage.input_tokens

number

Yes

Input tokens counted before detection.

usage.credits_charged

number

Yes

Credits deducted (always 1 credit).

credits_remaining

number

Yes

Remaining API credits after billing.

Response Example

{
  "code": 200,
  "message": "Success",
  "data": {
    "request_id": "req_1711863600789_f0ab19ce2",
    "is_ai_generated": true,
    "ai_probability": 84.2,
    "human_probability": 15.8,
    "confidence": 0.93,
    "usage": {
      "input_tokens": 411,
      "credits_charged": 1
    },
    "credits_remaining": 95
  },
  "timestamp": 1711863600832
}

Error Handling

Business failures are returned with an HTTP status code and the same outer response wrapper shape. Handle the HTTP status code first, then read message for user-facing details.

400

Bad Request

Validation failure, text too short, content too long, or unsupported payload shape.

401

Unauthorized

Missing or invalid X-API-Key.

402

Payment Required

Insufficient API credits for the requested operation.

429

Too Many Requests

Mapped rate-limit error. Can also surface when upstream polling is rate-limited.

Common error messages you should expect include text_too_short, text_too_long, content_too_long, topic_required, content_required, insufficient_credits, upstream_rate_limited, and internal_error.

Ready to generate your first request?

Create an API key, check your remaining balance, and start with the humanize example above. Every endpoint example on this page is designed to be copied and adapted directly.

Open Dashboard Jump to First Endpoint