Text and documents

PP-OCRv6 API: OCR for screenshots and image regions

PublicModel licence: Apache 2.0

PP-OCRv6 reads text from screenshots, photos of labels and image regions, and returns text lines with confidence, line boxes and word boxes.

Read the docsGet an API keyComing soonAPI reference

Price

Billed item	Unit	Credits
Text recognition One image per request.	per image	0.25 ≈ US$0.00025 top-up

Included in the public pool: 100 calls per day after sign-up. Prices are in credits. Failed calls cost nothing.

Sample results

Recorded responses of the API on sample images.

App settings screenshot: 35 recognised text lines with line boxes. — **App settings screenshot**1 image35 lines

Receipt: 44 recognised text lines with line boxes. — **Receipt**1 image44 lines

Code editor crop: 39 recognised text lines with line boxes. — **Code editor crop**1 image39 lines

Request and response

POST https://api.stack.foldbun.com/v1/ocr with a Bearer API key.

API reference

curlPythonRequest

# macOS: replace "base64 -w0 settings.webp" with "base64 -i settings.webp"
curl https://api.stack.foldbun.com/v1/ocr \
  -H "Authorization: Bearer $FOLDBUN_API_KEY" \
  -H "Content-Type: application/json" \
  -d @- <<EOF
{"model": "ppocr-v6", "image": {"base64": "$(base64 -w0 settings.webp)"}}
EOF

import base64, os, requests

with open("settings.webp", "rb") as f:
    image = base64.b64encode(f.read()).decode()

r = requests.post(
    "https://api.stack.foldbun.com/v1/ocr",
    headers={"Authorization": f"Bearer {os.environ['FOLDBUN_API_KEY']}"},
    json={"model": "ppocr-v6",
          "image": {"base64": image}},
    timeout=60,
)
r.raise_for_status()
print(r.json())

Response (shortened)

{
  "id": "call_01a1913d-55cd-7bee-ba86-bbbef2aa9bae",
  "model": "ppocr-v6",
  "image": {"height": 760, "id": "img_2dedd071…22e6", "width": 1100},
  "lines": [
    {
      "text": "Northwind",
      "confidence": 1,
      "bbox": [55, 22, 161, 47],
      "quad": [
        [55, 22],
        [161, 22],
        [161, 47],
        [55, 47]
      ],
      "words": [
        {
          "text": "Northwind",
          "confidence": 1,
          "bbox": [67, 22, 154, 47],
          "quad": [
            [67, 22],
            [154, 22],
            [154, 47],
            [67, 47]
          ]
        }
      ]
    }
    … 34 more
  ],
  "cost": 0.25,
  "balance": 1234.31
}

Request fields

Field	Description
image	`{"base64": "…"}` or `{"id": "img_…"}` from an earlier response.
options.minConfidence	Drop lines below this confidence, 0–1, default 0.5.
options.words	Include word-level boxes, default `true`.

Pricing examples

Call	Credits
1 image 0.25	0.25 ≈ US$0.00025 top-up
1,000 images 1,000 × 0.25	250 ≈ US$0.25 top-up

Limits

Images: JPEG, PNG, static WebP; up to 20 MiB and 40 megapixels
Coordinates: Pixels of the original image with EXIF orientation applied; masks can be smaller and carry their own size
Images per call: 1
Rate limit: Depends on the account tier; see GET /v1/models

Frequently asked questions

How is PP-OCRv6 billed?

Each image costs 0.25 credits, whatever the number of lines. The model is in the public pool: 100 calls per day at no charge after sign-up. Failed calls cost nothing.

What is the difference between ppocr-v6 and paddleocr-vl?

ppocr-v6 reads text lines from screenshots and image regions and costs less. paddleocr-vl parses a whole document page into ordered blocks, tables and formulas and returns Markdown, at 2.0 credits per page.

Are word boxes returned?

Yes, by default. Each line carries its text, confidence, box, quadrilateral and word boxes. Set options.words to false to omit the word boxes.

Related models

Models and pricing

Foldbun Stack — pay-per-use vision model API

Call PP-OCRv6 through the REST API; the docs list request and response formats.

Read the docsGet an API keyComing soon