canonical: https://jentic.com/apis/openalpr.com/openalpr # OpenALPR CarCheck API The OpenALPR CarCheck API reads license plates and vehicle details from images of cars, returning the plate text along with candidate alternatives and confidence scores. Images can be sent as an upload, as raw bytes, or by URL, and an optional flag adds vehicle make, model, color, and body-type recognition on top of the plate read. A country parameter tunes the recognizer to regional plate formats, and a configuration endpoint reports what the service supports. Developers building parking, tolling, security, and fleet applications use it to turn a camera frame into structured plate and vehicle data. ## For AI agents Recognize license plates and optional vehicle make, model, and color from an uploaded image, raw bytes, or an image URL with OpenALPR CarCheck. ## Scope Does not look up registration records or owner details from a plate. Use it to recognize plates and optional vehicle make, model, color, and body type from images only. ## Capabilities - Recognize a license plate from an uploaded image - Recognize a plate from raw image bytes - Recognize a plate from an image URL - Add vehicle make, model, color, and body-type recognition to a read - Tune recognition to a country's plate formats - Read the service configuration and supported options ## Use cases ### Parking and Access Control A gate or parking agent sends a camera frame to POST /recognize and reads the returned plate to decide whether to open a barrier. Passing the `country` parameter tunes the recognizer to local plate formats so reads stay accurate for the region. Example prompt: Read the license plate from the entry-camera image and return the top candidate with its confidence ### Vehicle Make and Model Lookup With the `recognize_vehicle` flag set, an agent calls POST /recognize to get the make, model, color, and body type alongside the plate. This enriches a security or fleet record with vehicle attributes from a single image. Example prompt: Recognize the plate and vehicle make, model, and color from a car photo for a fleet record ### Recognition from a Hosted Image URL When the image already lives at a URL, an agent calls POST /recognize_url with `image_url` so the service fetches and processes it directly. This avoids downloading and re-uploading frames stored in object storage. Example prompt: Recognize the plate from an image URL in object storage and return the result ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | `/recognize` | Recognize a plate from an uploaded image | | POST | `/recognize_bytes` | Recognize a plate from raw image bytes | | POST | `/recognize_url` | Recognize a plate from an image URL | | GET | `/config` | Read the service configuration | ## Key resources - **Recognition** — Plate and vehicle recognition from image, bytes, or URL - **Configuration** — Reported service configuration and supported options ## Why Jentic - **Setup:** Calling the OpenALPR CarCheck API by hand means attaching the `secret_key` query parameter, choosing the right recognition endpoint for your image source, and parsing candidate results yourself. Through Jentic you install once, import the API from the API Directory, store your secret key once, and your agent recognizes plates. - **Permission scoping:** OpenALPR takes its inputs as query parameters rather than in the path, so scope by operation: allow an agent the recognition endpoints it needs, for example image upload, and leave URL-based recognition out of the allowed set until you add it. - **Credential handling:** Your OpenALPR secret key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'recognize a license plate' or 'read vehicle make and model', and Jentic returns the matching OpenALPR operation with its input schema so the agent calls the right endpoint. ## Related APIs - **DeepAI API** — DeepAI offers broad image models; OpenALPR is specialized for license plate and vehicle recognition. - **Clarifai API** — Clarifai classifies and detects general image content; OpenALPR focuses on plates and vehicles. - **Imagga Image Recognition API** — Imagga tags and categorizes images; OpenALPR extracts the specific plate and vehicle details. ## FAQ ### What does the OpenALPR CarCheck API do? It reads license plates from car images and returns the plate text with candidate alternatives and confidence scores. With a flag set, it also recognizes vehicle make, model, color, and body type. Images can be sent as an upload, as raw bytes, or by URL. ### What authentication does the OpenALPR CarCheck API use? Each recognition call includes a `secret_key` query parameter that identifies your account. Through Jentic that key is stored once, encrypted, by your own instance and supplied at execution time, so the agent never reads the raw value. ### How do I recognize a plate from an image? Send the image to POST /recognize with your `secret_key` and a `country` value. To read vehicle attributes as well, set the `recognize_vehicle` flag, and use POST /recognize_url when the image is already hosted at a URL. ### Can it identify the vehicle, not just the plate? Yes. Setting `recognize_vehicle` on a recognition call adds make, model, color, and body-type predictions alongside the plate read, which is useful for fleet and security records. ### What are the rate limits for the OpenALPR CarCheck API? The OpenAPI specification does not declare rate limits. Send frames at a sensible cadence and back off on error responses; confirm the limits tied to your plan in the OpenALPR documentation. ### Do I need an MCP server to use the OpenALPR CarCheck API with an agent? You do not. Jentic connects the API to your agent directly from the API Directory: import it, store your secret key once, and the agent calls the recognition operations without a separate server to run. ### Can I limit what my agent is allowed to do with the OpenALPR CarCheck API? Yes. Because you run Jentic One yourself, you choose the allowed operations: an agent can be permitted to call recognition from an image while URL-based recognition stays excluded, and the secret key stays with your instance and is applied at call time.