👁️ OCR (Optical Character Recognition)

This API performs OCR (Optical Character Recognition) on scanned documents. It can either translate scanned files directly or extract the recognized HTML content without translation.

📋 Basic Information

Method: POST
Path: https://api.bluente.com/api/20250924/blu_translate/translate
Authentication: Required — Authorization: Bearer $BLUENTE_API_KEY

📝 Request Parameters

Query String

Field	Type	Required	Description
`engine`	int	Yes	Translation engine: `3 = LLM`, `4 = LLM Pro`. ⚠️Warning: The engine must be consistent with the task Upload File.
`scanned`	int	Yes	Must be `1`. Indicates the file is scanned and requires OCR
`only_extract`	int	Yes	`1 = Extract scanned HTML only (no translation)`. Must be used with `scanned=1`. The HTML file can be downloaded via Download File API

Request Body

Field	Type	Required	Description
`id`	int	Yes	Task ID returned from Upload File API
`action`	string	Yes	`"start"` to start OCR/translation, `"cancel"` to cancel the task
`from`	string	Yes	Source language code
`to`	string	Yes	Target language code
`namespace`	string	No	Use a specific namespace for glossary. If not set, the entire glossary is used

Example (OCR Translation)

curl -X POST "https://api.bluente.com/api/20250924/blu_translate/translate?engine=3&scanned=1" \
     -H "Authorization: Bearer $BLUENTE_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{
           "id": 1,
           "action": "start",
           "from": "en",
           "to": "zh",
           "namespace": "Bluente Partner"
         }'

fetch("https://api.bluente.com/api/20250924/blu_translate/translate?engine=3&scanned=1", {
  method: "POST",
  headers: {
    "Authorization": "Bearer $BLUENTE_API_KEY"
  },
  body: JSON.stringify({
    id: 1,
    action: "start",
    from: "en",
    to: "zh",
    namespace: "Bluente Partner"
  })
})

Example (OCR Only Extract)

curl -X POST "https://api.bluente.com/api/20250924/blu_translate/translate?engine=3&scanned=1&only_extract=1" \
     -H "Authorization: Bearer $BLUENTE_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{
           "id": 1,
           "action": "start",
           "from": "en",
           "to": "zh"
         }'

fetch("https://api.bluente.com/api/20250924/blu_translate/translate?engine=3&scanned=1&only_extract=1", {
  method: "POST",
  headers: {
    "Authorization": "Bearer $BLUENTE_API_KEY"
  },
  body: JSON.stringify({
    id: 1,
    action: "start",
    from: "en",
    to: "zh"
  })
})

✅ Success Response Example

{
  "message": "success",
  "code": 0
}

❌ Error Response Example

{
  "message": "invalid scanned parameter",
  "code": 400
}

📌 Notes

scanned=1 is mandatory for OCR requests.
If only_extract=1, the system will only generate an OCR HTML file without translation.
The extracted HTML can be downloaded via the Download File API.
Use action=start to begin OCR or action=cancel to stop it.
After starting OCR, check progress using the Get Translation Status API.

Translate file Download file