OCR API — Smallapis

4 endpoints

Click one to see its parameters and a ready-to-paste curl call.

POST /v1/ocr/image Read the text in an image

Read the text in an image

Returns: json

Parameter	Type	Description
`file`	`file`	Image to read.
`url`	`string`	Public URL instead of an upload.
`lang`	`string`	eng, fra, deu, spa, ita, por, nld. Combine with "+". Default: `eng`
`words`	`boolean`	Also return per-word bounding boxes and confidence. Default: `false`
`psm`	`integer`	Tesseract page segmentation mode, 0-13. Default: `3`

curl -X POST "https://173-249-18-142.sslip.io/v1/ocr/image" \
  -H "X-API-Key: $SMALLAPIS_KEY" \
  -F "file=@input.jpg" \
  -F "lang=eng" \
  -F "words=false" \
  -F "psm=3"

POST /v1/ocr/pdf Read the text in a scanned PDF

Uses the embedded text layer when there is a usable one, and only falls back to OCR when the pages are images.

Returns: json

Parameter	Type	Description
`file`	`file`	PDF to read.
`lang`	`string`	OCR language. Default: `eng`
`first`	`integer`	First page, 1-based. Default: `1`
`last`	`integer`	Last page. At most 20 pages per call.
`force`	`boolean`	OCR even when a text layer exists. Default: `false`

curl -X POST "https://173-249-18-142.sslip.io/v1/ocr/pdf" \
  -H "X-API-Key: $SMALLAPIS_KEY" \
  -F "file=@input.pdf" \
  -F "lang=eng" \
  -F "first=1" \
  -F "force=false"

POST /v1/ocr/searchable-pdf Add an invisible text layer to a scan

Takes an image or a scanned PDF and returns a PDF you can search and copy from.

Returns: binary pdf

Parameter	Type	Description
`file`	`file`	Image or PDF to process.
`lang`	`string`	OCR language. Default: `eng`

curl -X POST "https://173-249-18-142.sslip.io/v1/ocr/searchable-pdf" \
  -H "X-API-Key: $SMALLAPIS_KEY" \
  -F "file=@input.pdf" \
  -F "lang=eng" \
  --output result.bin

GET /v1/ocr/languages List the installed OCR languages

List the installed OCR languages

Returns: json

No parameters.

curl -H "X-API-Key: $SMALLAPIS_KEY" \
  "https://173-249-18-142.sslip.io/v1/ocr/languages" --output out.bin