Deep learning–powered OCR engine with high accuracy. Supports letters, numbers, mixed characters, and arithmetic captchas. Millisecond response, ready out of the box.
GPU-accelerated inference, avg. response < 200ms
All calls free for now, no rate limits
Standard RESTful API, no SDK required
Supported types
Covers mainstream captcha types — sample recognition results below




Quick start
Pick your language and send a POST request
import requests
url = "https://ocr-api.pixmagic.app/recognize"
files = {'file': open('captcha.png', 'rb')}
response = requests.post(url, files=files)
print(response.json())
# => {"captcha_text": "BA3K"}curl -X POST https://ocr-api.pixmagic.app/recognize \
-F "file=@captcha.png"const formData = new FormData();
formData.append('file', captchaFile);
fetch('https://ocr-api.pixmagic.app/recognize', {
method: 'POST',
body: formData,
})
.then(r => r.json())
.then(data => console.log(data.captcha_text));No sign-up, no API key — just call the endpoint. We may introduce paid plans later, but everything is fully open today.
Get started