PaddleOCR-VL
Free.ai (self-hosted)
·
ocr
·
~300 tokens per page
PaddleOCR-VL is an OCR model built by PaddlePaddle. Strongest at Fast multilingual image / scan to Markdown; tables, formulas, charts; 109 languages.. Self-inotarisirwa pa Free.ai GPUs - inofamba zvakasununguka pazuva rako re token pool (300 tokens peji). Yakaburitswa pasi pe Apache 2.0 — kushandiswa kwekutengesa kwakabvumirwa pa Free.ai.
Usati washandisa
OpenAI-inowirirana REST API. Kugadzira chiratidzo uye kukumbira iyi model mumasekondi.
curl -X POST https://api.free.ai/v1/ocr/ \
-H "Authorization: Bearer sk-free-..." \
-H "Content-Type: application/json" \
-d '{"model":"paddleocr-vl","prompt":"your prompt here"}'Zvimwe zvinobvunzwa kakawanda
PaddleOCR-VL inotora mazita emashoko kubva kumapikicha uye maPDF. Upload a scanned page, a screenshot, kana a photo of a document and PaddleOCR-VL returns the text — preserving line breaks and (on capable engines) tables, formulas, and structure.
PaddleOCR-VL inobata mazhinji emazita ezvinyorwa — Latin, Cyrillic, CJK, Arabic, Devanagari, nezvimwe. /ocr/ landing ine yakazara runyorwa rwezvinyorwa zvechinyakare.
Yes for clean printed-style handwriting; cursive and shorthand are harder. TrOCR and modern transformer-based engines are noticeably better than legacy Tesseract on handwriting.
Layout-aware engines (GOT-OCR, Marker, Florence-2) idzoreredza Markdown/HTML mapepanhau uye LaTeX for formulas. Plain-text engines idzoreredza mapepanhau pasina hunhu — tarisa izvi model's output mode.
PaddleOCR-VL inofamba paGPU yedu kubva kune yako yezuva nezuva yemahara pool; $ 5 → 200,000 yakabhadharwa ma tokens mushure me. ~ 300 ma tokens papeji.
PNG, JPG, WebP, HEIC, BMP, pamwe nePDF yepakutanga uye yepamberi kusvika pa50 MB. Zvinyorwa zvakasviba kana zvakasviba zvinogadziriswa otomatiki mushure mekubvisa.
Kunyatsonzwisisa kwechiratidzo ndekwe> 98% papepa rakanaka rakadhindwa, kudonha pamafoto ane simba rakanyanya, kusvibiswa kwescanner, kana mafonti asina kunaka.
Yeah — /batch/ inogamuchira fomati yemafaira emafaira / PDFs. Yese yekubvisa inosvika mu /account/?tab=history. Kuti uchengetedze fomati yemugwagwa, shandisa API.
Yeah — POST yako faira kuti /v1/ocr/ with model="PaddleOCR-VL". Returns JSON with text + (kana yakatsigirwa) layout + word-level bounding boxes. /api/ ine yakazara reference.
Self-hosted mamodheru anochengeta faira rako paGPU yedu; premium inoenda kuburikidza neDPA. Mapepa anogadziriswa mushure mekugovera-window. Hatidzidzisi pane ako mavhoti.
Yeah — Free.ai inobvumidza kushandiswa kwezvigadzirwa zvemapepa. Unofanira kunge uine kodzero dzekushandisa chinyorwa chemutauro wechigadzirwa.
1-5 masekondi papeji rimwe chete. Layout-aware injini (mapepa, math) zvinotora nguva yakareba.