How it works

Kinoux exposes each model through a standard endpoint. You send input data, and the model returns a structured response.

The API is designed to be simple and consistent, making it easy to integrate into tools like Zapier or Make.

01

Create your API key in your Kinoux dashboard.

02

Select a model and send your input data to the /predict endpoint.

03

Use the model output directly in your workflow or application.

Explore available models

Use the Models API to browse active models or retrieve details for a specific model before sending a prediction request.

List available models

GET /api/v1/models

Retrieve the list of active models available on your account, including their endpoints.

Example response

{
	"data": [
		{
			"code": "FR_SPAM_HAM",
			"name": "French Spam / Ham Detection",
			"slug": "french-spam-ham-detection",
			"call_url": "/api/v1/models/french-spam-ham-detection/predict",
			"is_active": true
		}
	]
}

Get model details

GET /api/v1/models/{slug}

Retrieve detailed information for a specific model, including its endpoint and example input.

Example response

{
	"data": {
		"code": "FRENCH_SPAM_TYPE_DETECTION",
		"name": "French Spam Type Detection",
		"slug": "french-spam-type-detection",
		"call_url": "/api/v1/models/french-spam-type-detection/predict",
		"example_payload": {
			"text": "Cliquez ici pour récupérer vos gains."
		},
		"is_active": true
	}
}

Run a model

POST /api/v1/models/{slug}/predict

Send input data to a model and receive structured predictions in real time.

Auth Authorization: Bearer <api_key>
Notes One request processes one input. Batch requests are not supported.

Example request

{
	"input": {
		"text": "Offre gratuite, cliquez ici"
	}
}

Example response

{
	"request_id": "e922f5a6-45dd-45af-85f7-9f11320c4e7f",
	"model": {
		"code": "FRENCH_SPAM_HAM_DETECTION",
		"name": "French Spam/Ham Detection"
	},
	"output": {
		"label": "spam",
		"score": 0.97
	},
	"meta": {
		"status": 200,
		"latency_ms": 143
	}
}

Notes and common errors

The Models API uses API key authentication and processes one input per request.
Most issues come from authentication, invalid model slugs, or quota limits.

Authentication

Use your API key in the Authorization header for every request.

Example header

Authorization: Bearer <api_key>

Request limits

Batch payloads are not supported on this endpoint. Send one input per request.

Notes

  • One request processes one input
  • Batch requests are not supported
  • Quota limits may apply depending on your plan

Common errors

Common errors returned by the Models API include:

  • 401 — Invalid or missing API key
  • 403 — Email not verified or API scope forbidden
  • 404 — Model not found
  • 422 — Batch payload not supported
  • 429 — Quota exceeded

Start using the API

Create an account, get your API key, and start sending requests in minutes.

Try for free