PaddleOCR — OCR AI Model

PaddleOCR is tuned for Korean printed text and needs no GPU at all, which makes it the one to run over a large pile of documents. For layout, tables or handwriting, use GLM-OCR or DeepSeek OCR instead.

Try in Studio

Details

  • CategoryOCR
  • Year2025
  • LicenseApache-2.0

Compliance & Provenance

  • ProviderOpen-source · Specialized
  • EU AI Act RiskMinimal Risk
  • Art. 50 TransparencyNot applicable

Inputs & Outputs

  • ImageInput · image

    Source image containing text to recognize

  • TextOutput · string

    Detected texts from an image

  • Region DataOutput · string

    JSON: one record per detected line with its text, recognition score, 4-point polygon, and bounding box

Tags

  • ocr
  • text-recognition
  • korean
  • text-extraction
  • no-gpu
  • fast

Alternatives in Text Recognition(OCR)

  • DeepSeekOCR

    Multi-language + handwriting OCR. 30-60s.

  • GLM-OCR

    Document OCR with layout/table/formula. 10-40s.

Resources