API Documentation — AI Text Humanization Engine

Quick Start

curl -X POST https://texthumanize.link/api/humanize \
  -H "Content-Type: application/json" \
  -d '{"text": "Your AI text here", "lang": "en", "intensity": 60}'

import requests

response = requests.post("https://texthumanize.link/api/humanize", json={
    "text": "Your AI text here",
    "lang": "en",
    "intensity": 60,
    "profile": "web"
})
data = response.json()
print(data["text"])

const response = await fetch("https://texthumanize.link/api/humanize", {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({ text: "Your AI text here", lang: "en", intensity: 60 })
});
const data = await response.json();
console.log(data.text);

$response = file_get_contents("https://texthumanize.link/api/humanize", false,
    stream_context_create(["http" => [
        "method" => "POST",
        "header" => "Content-Type: application/json",
        "content" => json_encode(["text" => "Your AI text here", "lang" => "en", "intensity" => 60])
    ]])
);
$data = json_decode($response, true);
echo $data["text"];

Authentication

All endpoints work without authentication up to 3 requests/day (guest mode). For unlimited access, register for a free account and create an API key from your dashboard.

Bearer Token

Include your API key in the Authorization header:

HTTP Header

Authorization: Bearer th_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

curl -X POST https://texthumanize.link/api/humanize \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer th_your_api_key_here" \
  -d '{"text": "Your AI text here", "lang": "en"}'

import requests

API_KEY = "th_your_api_key_here"

response = requests.post(
    "https://texthumanize.link/api/humanize",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={"text": "Your AI text here", "lang": "en", "intensity": 60}
)
data = response.json()
print(data["text"])

const API_KEY = "th_your_api_key_here";

const response = await fetch("https://texthumanize.link/api/humanize", {
    method: "POST",
    headers: {
        "Content-Type": "application/json",
        "Authorization": `Bearer ${API_KEY}`
    },
    body: JSON.stringify({ text: "Your AI text here", lang: "en" })
});
const data = await response.json();
console.log(data.text);

<?php
$apiKey = 'th_your_api_key_here';
$ch = curl_init('https://texthumanize.link/api/humanize');
curl_setopt_array($ch, [
    CURLOPT_RETURNTRANSFER => true,
    CURLOPT_POST => true,
    CURLOPT_HTTPHEADER => [
        'Content-Type: application/json',
        'Authorization: Bearer ' . $apiKey,
    ],
    CURLOPT_POSTFIELDS => json_encode(['text' => 'Your AI text here', 'lang' => 'en']),
]);
$response = json_decode(curl_exec($ch), true);
curl_close($ch);
echo $response['text'];

Rate Limits by Plan

Plan	API Keys	Quota	PAYG Overage
Guest (no account)	—	3 requests/day	—
Free	—	10 000 chars/day	—
Starter ($9/mo)	1 key	500 000 chars/mo	$0.001/1K chars
Pro ($19/mo)	5 keys	3 000 000 chars/mo	$0.001/1K chars
Business ($49/mo)	20 keys	Unlimited	—

Getting an API Key

Register for a free account (or sign in).
Upgrade to Starter or higher from Pricing.
Go to your Dashboard and click Create API Key.
Copy the key — it's shown only once at creation time.

You're browsing as a guest (3 free requests/day). Register free

Base URL & Rate Limits

https://texthumanize.link/api

Parameter	Value
Base URL	`https://texthumanize.link/api`
Protocol	HTTPS only
Content-Type	`application/json`
Rate Limit (guest)	3 requests/day (session-based, no registration needed)
Rate Limit (free plan)	10 000 chars/day, unlimited requests
Rate Limit (starter+)	Based on plan quota — Starter: 500K chars/mo, Pro: 3M chars/mo, Business: unlimited
CORS	All origins allowed (`*`)
Authentication	Optional Bearer token (`Authorization: Bearer th_…`) — required for higher limits and paid plans
Max Text Length	No limit

Endpoints Overview

Method	Endpoint	Description
POST	`/api/humanize`	Humanize AI-generated text
POST	`/api/detect`	Detect AI-generated text
POST	`/api/paraphrase`	Paraphrase text
POST	`/api/tone/analyze`	Analyze text tone
POST	`/api/tone/adjust`	Adjust text tone
POST	`/api/watermark/detect`	Detect watermarks in text
POST	`/api/watermark/clean`	Remove watermarks from text
POST	`/api/spin`	Spin / rewrite text content
POST	`/api/coherence`	Analyze text coherence
POST	`/api/readability`	Analyze readability metrics
POST	`/api/grammar/check`	Check grammar issues
POST	`/api/grammar/fix`	Auto-fix grammar issues
POST	`/api/perplexity`	Calculate text perplexity
POST	`/api/health`	Content health score
POST	`/api/plagiarism`	Check originality / plagiarism
GET	`/api/ping`	Health check

Error Handling

All errors return JSON with "ok": false and an "error" message.

HTTP Code	Meaning	Example
`200`	Success	`{"ok": true, ...}`
`400`	Bad Request — missing or invalid parameters	`{"ok": false, "error": "Text is required"}`
`404`	Not Found — unknown endpoint	`{"ok": false, "error": "Unknown action: foo"}`
`405`	Method Not Allowed	`{"ok": false, "error": "Method not allowed"}`
`401`	Unauthorized — invalid or missing API key	`{"ok": false, "error": "Invalid API key"}`
`429`	Rate Limited — guest daily limit or plan quota exhausted	`{"ok": false, "error": "guest_limit_reached", "daily_limit": 3}`
`500`	Internal Server Error	`{"ok": false, "error": "..."}`

Endpoint Details

POST /api/humanize Humanize AI-generated text

Transforms AI-generated text into more natural, human-like content. Supports multiple profiles and intensity levels. For texts longer than 3 000 characters, processing is done in chunks automatically.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to humanize
`lang`	string	No	`"en"`	Language code: en, uk, ru, de, fr, es, pl, pt, it
`intensity`	integer	No	`60`	Humanization intensity (10–100). Higher = more changes
`profile`	string	No	`"web"`	Profile: general, chat, web, seo, docs, formal

Response Fields

Field	Type	Description
`ok`	boolean	Always `true` on success
`text`	string	The humanized text
`lang`	string	Language used
`profile`	string	Profile used
`intensity`	integer	Intensity used
`change_ratio`	float	Ratio of text changed (0.0–1.0)
`changes`	array	List of individual changes made
`metrics_before`	object	AI detection metrics before humanization
`metrics_after`	object	AI detection metrics after humanization
`elapsed_ms`	float	Processing time in milliseconds

Example

Request

curl -X POST https://texthumanize.link/api/humanize \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Machine learning algorithms utilize data to improve performance on specified tasks.",
    "lang": "en",
    "intensity": 70,
    "profile": "web"
  }'

Response

{
  "ok": true,
  "text": "Machine learning algorithms use data to get better at specific tasks.",
  "lang": "en",
  "profile": "web",
  "intensity": 70,
  "change_ratio": 0.31,
  "changes": [
    {"original": "utilize", "replacement": "use", "rule": "simplify_verb"},
    {"original": "improve performance on specified", "replacement": "get better at specific", "rule": "natural_phrasing"}
  ],
  "metrics_before": {"artificiality_score": 45.2, "avg_sentence_length": 11, "burstiness_score": 0.5},
  "metrics_after": {"artificiality_score": 18.7, "avg_sentence_length": 11, "burstiness_score": 0.5},
  "elapsed_ms": 12.3
}

POST /api/detect Detect AI-generated text

Analyzes text to determine the probability that it was generated by AI. Returns multiple detection signals including artificiality score, burstiness, sentence patterns, and per-sentence analysis.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to analyze
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`artificiality_score`	float	Overall AI probability (0–100)
`avg_sentence_length`	float	Average words per sentence
`bureaucratic_ratio`	float	Ratio of formal/bureaucratic words
`connector_ratio`	float	Ratio of transitional connectors
`repetition_score`	float	Word repetition metric
`burstiness_score`	float	Sentence length variation
`verdict`	string	"Human", "Mixed", or "AI"
`ai_detector`	object	Detailed detector output (aiProbability, confidence, verdict, entropyScore)
`sentences`	array	Per-sentence breakdown
`elapsed_ms`	float	Processing time

Example

Request

curl -X POST https://texthumanize.link/api/detect \
  -H "Content-Type: application/json" \
  -d '{"text": "Neural networks are a subset of machine learning.", "lang": "en"}'

Response

{
  "ok": true,
  "artificiality_score": 38.2,
  "avg_sentence_length": 9,
  "bureaucratic_ratio": 0.056,
  "connector_ratio": 0,
  "burstiness_score": 0.11,
  "verdict": "Mixed",
  "ai_detector": {"aiProbability": 0, "confidence": 0.1, "verdict": "unknown"},
  "elapsed_ms": 2.6
}

POST /api/paraphrase Paraphrase text

Rewrites text using synonyms and alternative phrasing while preserving meaning. Useful for avoiding duplicate content.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to paraphrase
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`text`	string	The paraphrased text
`change_ratio`	float	Ratio of words changed
`changes`	array	List of specific changes
`elapsed_ms`	float	Processing time

Example

Request

curl -X POST https://texthumanize.link/api/paraphrase \
  -H "Content-Type: application/json" \
  -d '{"text": "The research demonstrates significant improvements.", "lang": "en"}'

Response

{
  "ok": true,
  "text": "The study shows notable improvements.",
  "change_ratio": 0.28,
  "changes": [
    {"original": "research", "replacement": "study"},
    {"original": "demonstrates significant", "replacement": "shows notable"}
  ],
  "elapsed_ms": 8.1
}

POST /api/tone/analyze Analyze text tone

Analyzes the emotional tone and style of text. Returns detected tones with confidence scores.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to analyze
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`tones`	array	Detected tones with labels and scores
`dominant_tone`	string	The primary detected tone
`formality`	float	Formality level (0–1)
`elapsed_ms`	float	Processing time

Example

Response

{
  "ok": true,
  "tones": [{"label": "neutral", "score": 0.72}, {"label": "formal", "score": 0.61}],
  "dominant_tone": "neutral",
  "formality": 0.61,
  "elapsed_ms": 3.2
}

POST /api/tone/adjust Adjust text tone

Rewrites text to match a specified target tone while preserving original meaning.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to adjust
`lang`	string	No	`"en"`	Language code
`target_tone`	string	No	`"neutral"`	Target tone: casual, formal, friendly, professional, neutral

Response Fields

Field	Type	Description
`text`	string	The tone-adjusted text
`original_tone`	string	Detected original tone
`target_tone`	string	Requested target tone
`elapsed_ms`	float	Processing time

POST /api/watermark/detect Detect watermarks

Detects statistical watermark patterns that AI text generators embed in their output.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to analyze
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`has_watermark`	boolean	Whether watermark was detected
`confidence`	float	Detection confidence (0–1)
`patterns`	array	Detected watermark patterns
`elapsed_ms`	float	Processing time

POST /api/watermark/clean Clean watermarks from text

Removes detected watermark patterns from text by replacing watermarked word selections with alternatives.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to clean
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`text`	string	Cleaned text
`watermarks_removed`	integer	Number of watermarks removed
`elapsed_ms`	float	Processing time

POST /api/spin Spin text content

Rewrites text by replacing words and phrases with synonyms. Useful for creating content variations.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to spin
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`text`	string	The spun text
`changes`	array	List of word replacements
`elapsed_ms`	float	Processing time

POST /api/coherence Analyze text coherence

Evaluates how well sentences in the text flow together. Checks for logical transitions, topic consistency, and structural coherence.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to analyze
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`overall_score`	float	Overall coherence (0–1)
`sentence_pairs`	array	Coherence between adjacent sentences
`topic_consistency`	float	Topic consistency score
`elapsed_ms`	float	Processing time

POST /api/readability Analyze readability

Calculates multiple readability metrics including Flesch-Kincaid, Coleman-Liau, ARI, SMOG, Gunning Fog, and Dale-Chall.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to analyze
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`flesch_kincaid_grade`	float	Flesch-Kincaid grade level
`flesch_kincaid`	float	Alias for flesch_kincaid_grade
`coleman_liau_index`	float	Coleman-Liau Index
`ari`	float	Automated Readability Index
`smog_index`	float	SMOG Index
`gunning_fog`	float	Gunning Fog Index
`dale_chall`	float	Dale-Chall Readability Score
`avg_word_length`	float	Average word length in characters
`avg_syllables_per_word`	float	Average syllables per word
`avg_sentence_length`	float	Average words per sentence
`grade_level`	string	Human-readable grade level
`elapsed_ms`	float	Processing time

Example

Response

{
  "ok": true,
  "flesch_kincaid_grade": 4.79,
  "coleman_liau_index": 7.43,
  "ari": 4.84,
  "smog_index": 18.77,
  "gunning_fog": 6.22,
  "dale_chall": 7.05,
  "avg_word_length": 4.5,
  "avg_syllables_per_word": 1.36,
  "avg_sentence_length": 11,
  "flesch_kincaid": 4.79,
  "grade_level": "Grade 5",
  "elapsed_ms": 104.3
}

POST /api/grammar/check Check grammar

Checks text for grammar, spelling, and style issues. Returns individual issues with suggestions and also the auto-fixed text.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to check
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`issues`	array	Array of issues, each with: `rule`, `message`, `suggestion`, `severity`, `offset`, `length`
`text`	string	The auto-corrected text
`total_issues`	integer	Total number of issues found
`elapsed_ms`	float	Processing time

Example

Response

{
  "ok": true,
  "issues": [
    {
      "rule": "MORFOLOGIK_RULE_EN_US",
      "message": "Possible spelling mistake found.",
      "suggestion": "their",
      "severity": "error",
      "offset": 15,
      "length": 5
    }
  ],
  "text": "They went to their house after the long journey.",
  "total_issues": 1,
  "elapsed_ms": 89.4
}

POST /api/grammar/fix Fix grammar issues

Automatically fixes grammar and spelling issues in the text. Returns the corrected text directly.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to fix
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`text`	string	The corrected text
`elapsed_ms`	float	Processing time

POST /api/perplexity Calculate perplexity

Calculates the perplexity score of text — a measure of how predictable the text is to a language model. Lower perplexity suggests more natural, human-like text.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to analyze
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`score`	float	Perplexity score (lower = more natural)
`perplexity`	float	Raw perplexity value
`cross_entropy`	float	Cross-entropy of the text
`naturalness`	integer	Naturalness score (0–100)
`burstiness_score`	float	Perplexity variance across sentences
`verdict`	string	"natural", "mixed", or "artificial"
`sentence_details`	array	Per-sentence readability data (if available)
`elapsed_ms`	float	Processing time

Example

Response

{
  "ok": true,
  "cross_entropy": 18.99,
  "perplexity": 523415.54,
  "local_variance": 0,
  "burstiness_score": 0,
  "naturalness": 58,
  "verdict": "mixed",
  "score": 523415.54,
  "elapsed_ms": 71.5
}

POST /api/health Content health score

Provides a comprehensive content quality score combining readability, grammar, uniqueness, AI naturalness, and coherence into a single health grade.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to evaluate
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`score`	float	Overall health score (0–100)
`grade`	string	Letter grade: A, B, C, D, F
`components`	array	Array of components: `name`, `score`, `weight`, `details`
`elapsed_ms`	float	Processing time

Example

Response

{
  "ok": true,
  "score": 84.7,
  "grade": "B",
  "components": [
    {"name": "readability", "score": 100, "weight": 0.2, "details": "grade_level=10"},
    {"name": "grammar", "score": 100, "weight": 0.25, "details": "0 errors, 0 warnings"},
    {"name": "uniqueness", "score": 96.1, "weight": 0.2, "details": "vocab_richness=0.89"},
    {"name": "ai_naturalness", "score": 50, "weight": 0.2, "details": "verdict=human"},
    {"name": "coherence", "score": 70, "weight": 0.15, "details": "overall=0.00"}
  ],
  "elapsed_ms": 892.2
}

POST /api/plagiarism Check originality

Checks text originality using n-gram fingerprinting and self-similarity analysis. Returns an overall originality score and per-sentence breakdown.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to check

Response Fields

Field	Type	Description
`score`	integer	Originality score (0–100, higher = more original)
`originality_score`	integer	Same as score
`verdict`	string	"original", "partially_original", or "plagiarized"
`self_similarity`	float	Self-similarity ratio
`ngram_diversity`	float	N-gram uniqueness
`fingerprint_hash`	string	Content fingerprint
`sentence_originality`	array	Per-sentence originality scores
`elapsed_ms`	float	Processing time

Example

Response

{
  "ok": true,
  "originality_score": 80,
  "self_similarity": 0,
  "fingerprint_hash": "d791457b0ac29190",
  "ngram_diversity": 1,
  "sentence_originality": [
    {"sentence": "The quick brown fox...", "originality": 100, "is_original": true}
  ],
  "verdict": "original",
  "score": 80,
  "originality": 80,
  "elapsed_ms": 70.2
}

POST /api/quality Quality Score

Unified TextHumanize Quality Score (0–100 + letter grade) across seven dimensions: semantic similarity, naturalness, readability, AI-pattern resistance, watermark cleanliness, edit balance, and processing speed.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to score
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`score`	float	Composite quality score (0–1)
`score_100`	float	Composite score on a 0–100 scale
`grade`	string	Letter grade (A+ … F)
`verdict`	string	excellent \| good \| fair \| poor
`dimensions`	object	Per-dimension value, weight, and label
`recommendations`	array	Actionable next steps
`elapsed_ms`	float	Processing time

POST /api/audit AI & Watermark Audit

Combined AI-pattern and watermark risk report with highlighted spans and safe suggested actions. Describes internal risk signals; it is not a guarantee of bypassing any external detector.

Request Parameters

Parameter	Type	Required	Default	Description
`text`	string	Yes	—	The text to audit
`lang`	string	No	`"en"`	Language code

Response Fields

Field	Type	Description
`score`	float	Overall risk (max of AI and watermark risk, 0–1)
`ai`	object	Explainable AI-detection report (score, verdict, highlighted_spans, reasons)
`watermark`	object	Unicode + statistical watermark report (risk_score, has_watermarks, findings)
`suggested_actions`	array	Safe, neutral remediation steps
`elapsed_ms`	float	Processing time

GET /api/ping Health check

Simple health check endpoint to verify the API is running.

Example

Response

{
  "ok": true,
  "message": "pong",
  "version": "0.33.0",
  "timestamp": "2025-01-15T12:00:00+00:00"
}

SDKs & Libraries

Python pip install texthumanize TypeScript / Node.js npm install texthumanize PHP composer require ksanyok/text-humanize

PAYG Billing & Overages

On Starter and Pro plans, API usage exceeding your monthly quota is billed at $0.001 per 1 000 characters (pay-as-you-go). Business plan has unlimited usage with no overage charges.

Payments are processed via Monobank Acquiring (UAH). You can top up your balance or upgrade on the Pricing page.

Type	Rate	Example
PAYG overage (API)	$0.001 / 1K chars	100 000 chars = $0.10
Starter plan	$9 / month	500 000 chars included
Pro plan	$19 / month	3 000 000 chars included
Business plan	$49 / month	Unlimited