Image to Text (OCR) API

Authentication

All API requests require authentication using your API key. Send it via the x-api-key header with every request.

x-api-key: your_api_key_here

API endpoint

The API exposes a single synchronous endpoint. Send one image, get the extracted text and markdown back in the same response — there is no job ID and no polling.

Extract text from an image

POSThttps://api.corenexis.com/image-ocr/v1

How it works

The API is synchronous and processes exactly one image per request. The full lifecycle happens inside a single HTTP call.

1. Send the image — Send a POST request to /image-ocr/v1 with the image (file upload, URL, or base64) and an optional lang. The API reads the image, detects its MIME type, and measures its size.
2. Validation against your plan — Your API key, subscription, rate limit, and image size are checked against your plan. If lang is provided but unsupported, the request is rejected immediately — before any processing or quota check.
3. OCR runs instantly — The OCR engine reads the image. If no lang was given, the script is auto-detected (Latin → eng, Devanagari → hin, and so on) and the matching language is applied automatically.
4. Get text + markdown back — The response contains the plain text, the rebuilt markdown, character/word/line metrics, detected script, and processing time — all in one JSON body. One unit is deducted from your monthly quota only when the extraction succeeds.

Input modes — four ways to send an image

The API accepts the image in any of these formats. Use whichever fits your environment.

1. Direct file upload (multipart) — recommended

Upload the image as a multipart form field. The field name can be image, file, data, photo, img, upload, or attachment — all are accepted. Any common format works: JPG, PNG, WebP, GIF, BMP, TIFF, HEIC, AVIF.

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -F "image=@photo.jpg"

2. Public URL — image_url

Pass any public URL that returns image bytes. The image is validated by its actual content (MIME type), not by file extension — so URLs without an extension work too. For example, https://example.com/image that directly serves an image is accepted.

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -d '{"image_url":"https://cdn.example.com/receipt.png"}'

3. Google Drive share link — image_url

Paste a Google Drive share link directly. The API automatically detects Drive links (/file/d/ID/view, open?id=, uc?id=) and fetches the file. Make sure the file is shared with “Anyone with the link can view.” Dropbox links are also normalized to direct download.

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -d '{"image_url":"https://drive.google.com/file/d/1--EMd70IQHnLuVJXqXH...../view"}'

4. Base64-encoded image — image_base64

Send the image inline as base64. Useful when your client can’t perform multipart uploads (some no-code platforms). A data:image/png;base64,... prefix is accepted and stripped automatically.

IMG_B64=$(base64 -w0 photo.jpg)
curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -d "{\"image_base64\":\"$IMG_B64\"}"

Request parameters

Headers

Image source (one of)

Processing options

Code examples

Simple extraction (auto-detect language)

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -F "image=@photo.jpg"

const form = new FormData();
form.append('image', fileInput.files[0]);

const res = await fetch('https://api.corenexis.com/image-ocr/v1', {
  method: 'POST',
  headers: { 'x-api-key': 'your_api_key' },
  body: form
});
const data = await res.json();
console.log(data.data.text);

import requests

with open("photo.jpg", "rb") as f:
    res = requests.post(
        "https://api.corenexis.com/image-ocr/v1",
        headers={"x-api-key": "your_api_key"},
        files={"image": f}
    )
data = res.json()
print(data["data"]["text"])

$ch = curl_init("https://api.corenexis.com/image-ocr/v1");
curl_setopt_array($ch, [
    CURLOPT_POST => true,
    CURLOPT_RETURNTRANSFER => true,
    CURLOPT_HTTPHEADER => ["x-api-key: your_api_key"],
    CURLOPT_POSTFIELDS => ["image" => new CURLFile("photo.jpg")],
]);
$resp = json_decode(curl_exec($ch), true);
echo $resp["data"]["text"];

Force a specific language

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -F "image=@notice.png" \
  -F "lang=hin"

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -F "image=@bill.webp" \
  -F "lang=eng+hin"

import requests
with open("document.jpg", "rb") as f:
    res = requests.post(
        "https://api.corenexis.com/image-ocr/v1",
        headers={"x-api-key": "your_api_key"},
        files={"image": f},
        data={"lang": "fra"}
    )
print(res.json()["data"]["markdown"])

Image from URL / Google Drive

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -d '{"image_url":"https://cdn.example.com/receipt.png","lang":""}'

# URL has no .png/.jpg — detected by content, not extension
curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -d '{"image_url":"https://example.com/image"}'

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -d '{"image_url":"https://drive.google.com/file/d/1--EMd70IQHnLuVJXqXH...../view","lang":"eng+hin"}'

curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -F "image_url=https://cdn.example.com/screenshot.png"

Base64 image

IMG_B64=$(base64 -w0 photo.jpg)
curl -X POST "https://api.corenexis.com/image-ocr/v1" \
  -H "x-api-key: your_api_key" \
  -H "Content-Type: application/json" \
  -d "{\"image_base64\":\"$IMG_B64\"}"

import requests, base64

with open("photo.jpg", "rb") as f:
    b64 = base64.b64encode(f.read()).decode()

res = requests.post(
    "https://api.corenexis.com/image-ocr/v1",
    headers={"x-api-key": "your_api_key"},
    json={"image_base64": b64}
)
print(res.json()["data"]["text"])

Full extract + use result

import requests

API_KEY = "your_api_key"

with open("photo.jpg", "rb") as f:
    res = requests.post(
        "https://api.corenexis.com/image-ocr/v1",
        headers={"x-api-key": API_KEY},
        files={"image": f}
    ).json()

if not res["success"]:
    raise SystemExit(res["message"])

d = res["data"]
print("Detected lang:", d["lang"], "| auto:", d["lang_auto_detected"])
print("Words:", d["metrics"]["word_count"])
print("---- TEXT ----")
print(d["text"])
print("---- MARKDOWN ----")
print(d["markdown"])
print("Remaining quota:", res["usage"]["remaining"])

const API_KEY = 'your_api_key';

const form = new FormData();
form.append('image', fileInput.files[0]);

const res = await fetch('https://api.corenexis.com/image-ocr/v1', {
  method: 'POST',
  headers: { 'x-api-key': API_KEY },
  body: form
}).then(r => r.json());

if (!res.success) throw new Error(res.message);

const d = res.data;
console.log('Detected lang:', d.lang, '| auto:', d.lang_auto_detected);
console.log('Words:', d.metrics.word_count);
console.log(d.text);
console.log(d.markdown);
console.log('Remaining:', res.usage.remaining);

Response format

Because the API is synchronous, the extracted text comes back in the same response — there is no status to poll. A successful call returns data.status = "completed" with the full OCR output.

Success response

{
  "success": true,
  "plan": "starter",
  "data": {
    "status": "completed",
    "filename": "photo.jpg",
    "file_size": "242.5 KB",
    "image": { "width": 1280, "height": 720, "format": "jpeg" },
    "lang": "eng",
    "lang_auto_detected": true,
    "metrics": {
      "char_count": 1342,
      "char_count_no_spaces": 1098,
      "word_count": 233,
      "line_count": 41
    },
    "text": "Plain OCR text exactly as read...",
    "markdown": "## HEADING\n\nParagraph text rebuilt from wrapped lines..."
  },
  "usage": {
    "remaining": 498,
    "rate_limit": 20,
    "monthly_limit": 500
  }
}

Response fields

Supported languages

Leave lang blank for automatic detection. To force a language, pass its code. Combine multiple codes with + for multilingual images (e.g. eng+hin, chi_sim+eng). Multilingual and non-English codes require Starter plan or higher.

How auto-detect picks a language

The image’s script is detected, then the matching language is applied automatically:

Error codes

Every error response is consistent JSON: { success: false, code: "...", message: "..." }. Many errors include extra fields (param, provided, max_allowed) to help you self-correct without contacting support.

HTTP status reference

Example error responses

{
  "success": false,
  "code": "PARAM_LIMIT_EXCEEDED",
  "message": "Invalid value for 'max_file_size_bytes'. Allowed for your plan: 2097152 or less. Your request contains: 5242880. Please change 'max_file_size_bytes' or upgrade your plan.",
  "param": "max_file_size_bytes",
  "provided": 5242880,
  "op": "lte",
  "max_allowed": 2097152
}

{
  "success": false,
  "code": "PARAM_LIMIT_EXCEEDED",
  "message": "Feature 'lang' is not allowed on your current plan. Allowed for your plan: false. Your request contains: true. Please disable 'lang' or upgrade your plan.",
  "param": "lang",
  "provided": true,
  "allowed": false
}

{
  "success": false,
  "code": "INVALID_INPUT",
  "message": "Unsupported language code 'klingon'. See the docs for the supported list."
}

{
  "success": false,
  "code": "INVALID_INPUT",
  "message": "The provided file is not a valid image.",
  "detected_mime": "application/pdf"
}

{
  "success": false,
  "code": "QUOTA_EXCEEDED",
  "message": "Monthly quota reached. Upgrade your plan or wait for the next billing cycle."
}

{
  "success": false,
  "code": "RATE_LIMIT_EXCEEDED",
  "message": "Too many requests. Please wait before making another request."
}