Public REST API

API documentation

Extract English text from JPEG, PNG, and GIF images using Google Cloud Vision. Requests are synchronous and uploads or OCR results are never retained.

Base URL https://api.ocr.lelbaba.top

Quick start

No API key or authentication header is required.

Terminal

curl -X POST \
  -F "image=@test_image.jpg" \
  https://api.ocr.lelbaba.top/extract-text

Upload limits 10 MiB per image · 40 megapixels decoded · JPEG, PNG, or GIF

POST

/extract-text

Send one image as multipart/form-data.

Request

Field	Type	Required	Description
`image`	file	Yes	JPEG, PNG, or GIF image up to 10 MiB.

Query parameters

Parameter	Default	Description
`normalize`	`false`	Add cleaned text without changing the raw OCR result.
`metadata`	`false`	Add safe dimensions, byte size, color mode, and format.

Success response · 200

JSON

{
  "success": true,
  "text": "extracted text content here",
  "confidence": 0.95,
  "processing_time_ms": 1234
}

An image containing no readable text is still successful, with an empty text value and confidence: 0.0.

POST

/extract-text/batch

Send one to five repeated images fields. Each image may be 10 MiB; combined image data may not exceed 25 MiB. Results preserve input order.

Terminal

curl -X POST \
  -F "images=@first.jpg" \
  -F "images=@second.png" \
  https://api.ocr.lelbaba.top/extract-text/batch

A valid batch returns 200. Each item has its own success and status_code, so one invalid image does not discard successful results.

Response fields

Field	Type	Description
`success`	boolean	Whether OCR succeeded.
`text`	string	Raw text returned by Vision.
`confidence`	number	Confidence from 0.0 to 1.0.
`processing_time_ms`	integer	Server processing time.
`normalized_text`	string	Present when normalization is enabled.
`metadata`	object	Present when metadata is enabled.

Error responses

Application errors use one stable JSON envelope.

JSON

{
  "success": false,
  "error": {
    "code": "unsupported_image_format",
    "message": "Only JPEG, PNG, and GIF images are supported."
  },
  "processing_time_ms": 3
}

Status	Possible codes	Meaning
400	`malformed_request`, `invalid_batch`, `empty_upload`	Invalid multipart request, image count, or empty file.
413	`image_too_large`, `image_dimensions_too_large`, `batch_too_large`, `request_too_large`	A documented size or dimension limit was exceeded.
415	`unsupported_image_format`	Decoded image is not JPEG, PNG, or GIF.
422	`corrupt_image`	Image data is corrupt or unreadable.
503	`ocr_unavailable`	OCR dependency is temporarily unavailable.
504	`ocr_deadline_exceeded`	OCR exceeded its processing deadline.
500	`internal_error`	Unexpected sanitized server error.

GET

/health

Returns {"status":"ok"} without calling the OCR provider.