Detection and rewriting are worth more inside your pipeline than inside a browser tab you have to remember to open. TextSight ships a live REST API and native apps on both mobile platforms, so a score can be requested by a CMS before publish, by a script during a content review, or by a phone in the middle of a lecture. This page lists every surface, what it does, and — where a surface is not ready — says so instead of implying otherwise.
Each of these is live and reachable right now. Nothing in this section is a roadmap item.
Three text endpoints for detection, scoring and rewriting. Plain JSON in, plain JSON out, authenticated with a key. Available on Pro and above.
See the endpoints →TextSight: Detect & Humanize on the App Store. Detect and rewrite from a phone, signed in to the same account and the same quota as the web app.
View on the App Store →The same product on Google Play. One account across web, iOS and Android — usage counts against a single bucket rather than fragmenting per device.
View on Google Play →The full workspace at app.textsight.ai — detector, rewriter, Authenticity Score, history and the free tools. No install, nothing to configure.
Open the web app →The base URL is https://api.textsight.ai. GET /v2 is unauthenticated and returns this endpoint list, so you can confirm the service is reachable before writing a line of code.
| Endpoint | What it returns | Scope |
|---|---|---|
POST /v2/detect | The full AI-detection breakdown for a block of text. | scan |
POST /v2/score | A lightweight AI/human score on its own, when you do not need the breakdown. | scan |
POST /v2/rewrite | A single-pass humanized rewrite of the submitted text. | humanize |
A key presented to an endpoint outside its scope is rejected. That means a detection-only integration can hold a key that is structurally unable to spend rewrite quota — useful when the key lives in a CMS plugin or a shared CI runner.
Keys are in the format sk_live_… and are created from your account. Send the key either as a bearer token or in the dedicated header — both are accepted, so you can use whichever your HTTP client makes easier:
curl -X POST https://api.textsight.ai/v2/detect \
-H "Authorization: Bearer sk_live_your_key_here" \
-H "Content-Type: application/json" \
-d '{"text":"The text you want scored."}'Or swap the first header for x-api-key: sk_live_…. Nothing else changes.
API keys are issued on Pro and above. Free and Starter accounts cannot create a key, and the request is refused before it reaches the rate limiter. Per-minute limits are counted per account, not per key, so issuing extra keys does not raise your ceiling — it only lets you scope and revoke them independently.
| Plan | API keys | Requests per minute |
|---|---|---|
| Free | Not available | — |
| Starter | Not available | — |
| Pro | Yes | 60 |
| Business | Yes | 120 |
| Enterprise | Yes | 240 |
One caveat we will not bury: text detection is English-language, and no detector is 100% accurate. The API says so in its own discovery response. Treat a score as evidence to review, not a verdict to act on automatically. See the limitations page for where the detector is weakest.
We are not in the Zapier directory yet. That is worth stating plainly, because the workaround is genuinely good rather than a consolation prize.
The endpoints take a JSON body with a text field and a single auth header. There is nothing proprietary in the request shape, which means any tool that can send an authenticated HTTP POST can call TextSight — and every serious automation platform can:
Zapier. Use the built-in Webhooks by Zapier action, set the method to POST, the URL to the endpoint, and add the Authorization header. This gives you the same result a native app would, at the cost of one extra configuration step. Note that Webhooks by Zapier is a Premium action, so it needs a paid Zapier plan. Full step-by-step, with the fields to map →
Make, n8n, and Pipedream. All three have a generic HTTP module with header support and no premium gate on it. If you are choosing an automation tool specifically to run TextSight, these are the cheaper path.
Google Apps Script. UrlFetchApp.fetch() covers Docs, Sheets and Gmail, which is the fastest route to scoring a column of drafts in a spreadsheet.
Your CMS or CI. A pre-publish check in WordPress, a build step that scores changed Markdown files, a webhook on a form submission — all of these are a handful of lines against a plain JSON endpoint.
These are listed so you can plan around them. None of them is installable today, and we would rather lose a click than have you discover that after signing up.
Select-and-detect on any page, built on Manifest V3. The extension is built and the store listing is prepared, but it is not yet installable from the Chrome Web Store.
A Gutenberg sidebar that scores the post you are editing and rewrites flagged sentences in place. Built against the same API, not yet published to the WordPress.org directory.
Detection inside Docs, where the editor blocks the browser-extension approach. Built as an Apps Script add-on; the Marketplace listing has not gone live.
A native listing with detect and rewrite actions. Until it exists, the Webhooks by Zapier route does the same job against the live API.
Set it up with Webhooks →If one of these is blocking a decision, tell us which — it is the clearest signal we get about what to finish first.
Yes. The v2 REST API is live at https://api.textsight.ai, with POST /v2/detect, POST /v2/score and POST /v2/rewrite. Authenticate with an sk_live_ key as either a bearer token or the x-api-key header. GET /v2 is unauthenticated and returns the endpoint list.
Pro, Business and Enterprise. Free and Starter accounts cannot create keys. Per-minute limits are 60, 120 and 240 respectively, counted per account rather than per key.
Two. scan covers detect and score; humanize covers rewrite. A key used outside its scope is rejected, so you can issue a detection-only key that cannot spend rewrite quota.
Not yet — TextSight is not currently in the Zapier app directory. The REST API is the supported way to automate today, and it works from Webhooks by Zapier, Make, n8n, Pipedream and Apps Script. We would rather say this than list an integration that does not exist.
Text detection is English-language, and the API returns that caveat itself. Scores on non-English text are not reliable enough for us to present as a supported feature.
Yes. The endpoints take plain JSON and return plain JSON with no SDK, so anything that can POST with a header works — Apps Script, Make, n8n, Pipedream, Retool, Airtable scripting, and the generic webhook action in Zapier.
TextSight + Zapier — the five-step Webhooks setup, the response fields to map, and the limits that break automations.
API documentation — request and response shapes, error codes, and the worked examples for each endpoint.
AI humanizer API — what /v2/rewrite does to a draft, and how the rewrite is scored afterwards.
Detectors that offer an API — how the market compares on endpoints, limits and pricing.
Pricing — which tier issues keys, and what each one costs monthly and annually.
Detection limitations — where the detector is weakest, and why a score is evidence rather than proof.
API terms — the terms that govern programmatic use.
Create an account, upgrade to Pro or above, and issue a key. The first call is a single curl command — there is no SDK to install and no onboarding call to sit through.