One POST request tells you whether a piece of writing came from ChatGPT, Claude, Gemini or a human — and gives you a score for every single sentence, so you can show your users what was flagged instead of a number they have to trust.
1,000 free credits · no credit card · key is ready the moment you verify your email
/api/v1/detect/textcurl -X POST https://www.wasitaigenerated.com/api/v1/detect/text \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"content": "The implementation of machine learning..."
}'{
"isAI": true,
"confidence": 0.92,
"model": "tropa-3",
"patterns": ["AI patterns detected"],
"analysis": {
"likelihood": "AI-generated",
"reasoning": "Overall AI score: 92.3%"
},
"sentences": [
{
"text": "The implementation of machine learning...",
"isAI": true,
"confidence": 0.95,
"scores": { "ai": 0.95, "human": 0.05 }
},
{
"text": "I rewrote this part myself last night.",
"isAI": false,
"confidence": 0.81,
"scores": { "ai": 0.19, "human": 0.81 }
}
]
}Not a dashboard you have to send your users to — an endpoint your backend calls
From a passage to a closer look
Workflow illustrationThe study compares interviews from two communities. The findings suggest that local context shapes participation. The next section examines the evidence behind this conclusion.
Read the overall result
Start with the document-level AI assessment.
Inspect individual sentences
Use sentence-level scores to choose passages for a closer look.
Make an informed review
Check sources, drafting history and context before drawing a conclusion.
Every response includes a per-sentence breakdown. Highlight the flagged passages in your own UI — reviewers act on evidence, not on a percentage.
Our own detection model, retrained as new LLMs ship. Read the model write-up before you integrate.
JSON in, JSON out, Bearer auth. If your language can make an HTTP request, the integration is about fifteen lines.
1 credit per word — a 500-word essay costs 500 credits. No seat licences, no minimum contract, and one-time credits never expire.
Copy, paste your key, ship it
curl -X POST https://www.wasitaigenerated.com/api/v1/detect/text \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"content": "The implementation of machine learning..."
}'content — string, required. The text to analyse. text is accepted as an alias.isAI — boolean verdict for the whole document.confidence — 0–1, how strongly the verdict holds.model — the detection model that produced the result.sentences[] — per-sentence text, verdict and ai/human scores.Screen every submission on upload and surface the flagged sentences to the instructor inside your existing grading view.
Check freelance submissions against your editorial policy before they reach a human editor's queue.
Flag cover letters and written assessments that came out of a chatbot, without adding a step to your recruiters' workflow.
Catch mass-generated reviews and listing descriptions at write time, before they dilute the trust signal your buyers rely on.
Run every writer deliverable through the API as a QA gate and attach the score to the delivery record.
Support AI-disclosure requirements by logging a detection result alongside the content your organisation publishes.
Everything is a JSON body with an error field — handle these and you are done
1 credit per word. A 500-word essay costs 500 credits.
1,000
credits on email verification
≈ 1,000 words of testing. No card required.
$5 – $29
one-time, never expires
25,000 credits for $5, or 200,000 for $29 — roughly 200,000 words.
$299/mo
no per-word cost
Unlimited calls, whitelabel PDF reports, bulk processing.
A JSON object with isAI (boolean), confidence (0–1), the model used, a short analysis, and a sentences array. Each sentence carries its own isAI flag and ai/human scores, so you can highlight the flagged passages in your own interface instead of showing a single opaque percentage.
Text from the current generation of large language models — ChatGPT and GPT-4, Claude, Gemini, Llama, Mistral and the many wrappers built on them. Detection is based on statistical properties of the text itself, not on a list of known outputs, so new models are covered without waiting for an update.
Text detection costs 1 credit per word. Every verified account starts with 1,000 free credits — enough for roughly a thousand words of testing. After that, 25,000 credits are $5 and 200,000 credits are $29 as one-time purchases, or $299/month for unlimited API calls. Credits from one-time packs never expire.
Bearer token. Send Authorization: Bearer YOUR_API_KEY on every request. Your key is created automatically with your account and is visible in your dashboard. Keep it server-side — it spends your credits.
Send text in the content field as a JSON string. Very long documents are rejected with a 413 — split them into sections of a few thousand words and aggregate the scores. Longer text generally produces a more reliable verdict than a couple of sentences, because there is more signal to work with.
Yes. All paid plans include commercial use, so you can build the API into an LMS, a CMS, a marketplace or an applicant tracking system. For high volume or an SLA, contact us about an enterprise plan.
Text sent through the API is processed to produce a result and is not used to train our models. See the privacy policy for the full detail on retention.
As evidence, not as a verdict. Detection is probabilistic: heavily edited AI text can read as human, and some human writing is genuinely formulaic. Route high-confidence results to a human reviewer rather than auto-rejecting, and show the sentence-level breakdown so the reviewer can see what triggered the score.
Create an account, verify your email, and your API key is waiting in the dashboard.