REST API / AI detection

AI Detector API for review-ready signals

Send text to one HTTPS endpoint and bring AI writing analysis into your own product. Detector v3 returns structured data that your team can inspect, store, and explain in context.

Create an API key Read the docs

Use the key from a private server environment.

cURL request

curl --request POST \
  --url https://api.detecting-ai.com/api/detect/ \
  --header "Content-Type: application/json" \
  --header "Idempotency-Key: detector-review-001" \
  --header "X-API-Key: YOUR_API_KEY" \
  --data '{
    "text": "Paste the text you want to analyze.",
    "version": "v3"
  }'

POST requests to /api/detect/

X-API-Key authentication

Client Idempotency-Key for recoverable retries

Detector v3 for new builds

JSON request and response

Built for real review flows

A small API contract with room for careful judgment

The integration stays simple: your server sends the text, the detector analyzes it, and your interface decides how to present the result. That separation makes the workflow easier to test and keeps the final decision with a person.

One focused endpoint

Add a text string and detector version to a JSON request. There is no SDK to keep in sync for the basic integration.

Structured output

Read the detector details, version, and processed word count from the response instead of scraping a web page.

Human review stays visible

Use the signal to prioritize passages and cases. Do not turn one model score into an automatic accusation or rejection.

Request contract

Two fields are enough to start

Send UTF-8 text as JSON and choose the detector version deliberately. New integrations should use v3.

Field	Type	Required	What it does
text	string	Yes	The text you want the detector to analyze.
version	string	Yes	Detector version. Use v3 for a new integration.

Keep the full response during development

Start by logging the complete JSON response in a secure test environment. Once the contract is clear, map only the fields your product uses and retain the processed word count for usage records.

Example response

{
  "success": true,
  "data": {
    "details": {
      "result": "Detector analysis is returned here"
    },
    "version": "v3",
    "words_processed": 8
  }
}

Integration flow

From API key to a useful result

Build the request on your server first. Add product UI only after the happy path and failure cases are clear.

1
Prepare, save, and activate the key
Prepare a key in the dashboard, copy the one-time value into a server-side secret manager, confirm it is saved, and activate it. The full value is not shown again after activation.
2
Send a representative sample
POST realistic text with version v3 and persist a unique Idempotency-Key before sending. Reuse it only for an exact retry.
3
Present the evidence carefully
Translate the structured response into a clear review state, with space for a person to inspect the text and context.

Good fits

Useful wherever text already passes through a review queue

The API works best as one input to an existing decision process. It should help people decide where to look, not make the decision for them.

Detection is not authorship proof

Models can be wrong, especially with short, heavily edited, translated, or formulaic text. Keep consequential decisions out of fully automatic flows.

Editorial intake

Add an authorship signal to submitted drafts and route unusual cases to an editor for closer reading.

Education platforms

Return a review signal beside assignment context, revision history, citations, and the institution's own policy.

Internal moderation

Check supplied text inside a private workflow without sending your API key to the browser.

Batch triage

Process queued documents on the server and surface the cases that deserve a manual second look.

API FAQ

Practical answers before you ship

Start with a server-side request, log the response shape, then add the result to your existing review flow.

What endpoint does the AI Detector API use?+

Send a POST request to https://api.detecting-ai.com/api/detect/. The request body is JSON, authentication uses X-API-Key, and new integrations should persist a stable Idempotency-Key for every logical operation.

Which detector version should a new integration use?+

Use v3 for new integrations. The version is supplied in the JSON request body so the behavior is explicit and easy to test.

Can I call the API directly from browser JavaScript?+

Do not expose a private API key in browser code. Send requests from your server, an edge function, or another trusted environment and return only the result your interface needs.

Does an AI score prove who wrote the text?+

No. AI detection is a probabilistic signal. Use it to focus review on specific text, then consider drafts, sources, writing history, and the surrounding context.

Where can I check API usage and remaining words?+

The API dashboard shows the balance attached to your subscription. Each successful detector request reports the number of input words processed.

Build the first request

Make one test request before designing the workflow

Use Detector v3, keep the key on your server, and inspect the full JSON response with text that resembles your real input.

Open API docs Compare pricing