ISO 15924 · Tibt · abugida
Tibetan
The Tibetan writing system (ISO 15924 Tibt), a abugida: 54 languages use it, 207 Unicode characters. Best OCR accuracy: 81% Acc@5 (PaddleOCR-VL-1.5).
Languages written in Tibetan
- Amdo Tibetanadx
- Baltibft
- Brokpakesgt
- Bumthangkhakjz
- Bunanbfu
- Chalikhatgf
- Changthangcna
- Chocangacakhacgk
- Classical Tibetanxct
- Dakpakhadka
- Dolpodre
- Duhumbicvg
- Dur Brokkatbro
- Dzalakhadzl
- Dzongkhadzo
- Eastern Tamangtaj
- Gongdukgoe
- Gyalsumdo-Nubrikte
- Humlahut
- Iduclk
- Jadjda
- Kalaktang Monpakkf
- Khambakbg
- Khams Tibetankhg
- Khengkhaxkf
- Kurtokhaxkz
- Kutang Ghaleght
- Lakhalkh
- Layakhalya
- Leh Ladakhilbj
- Lepchalep
- Lowaloy
- Lunanakhaluk
- Manangenmm
- Mugommuk
- Olekhaole
- Pandebkj
- Pattanilae
- Purik-Sham-Nubraprx
- Sikkimesesip
- Solu-Khumbu Sherpaxsr
- Spiti Bhotispt
- Stod Bhotisbu
- Tibetanbod
- Tichurongtcn
- Tinanilbf
- Tshanglatsj
- Upper Mangdepneh
- Walunggeola
- Western Tamangtdg
- Zangskarizau
OCR results
- PaddleOCR-VL-1.5: Acc@5 81.0%, CER 3.8
- dots.mocr: Acc@5 78.0%, CER 4.6
- dots.ocr: Acc@5 76.0%, CER 5.0
- HunyuanOCR: Acc@5 71.0%, CER 5.6
- LightOnOCR-2-1B: Acc@5 21.0%, CER 25.7
- Gemini 3.1 Flash-Lite: Acc@5 19.0%, CER 15.3
- Qwen3-VL-8B: Acc@5 12.0%, CER 23.3
- FireRed-OCR: Acc@5 2.0%, CER 59.2
- DeepSeek-OCR-2: Acc@5 1.0%, CER 72.2
- GLM-OCR: Acc@5 0.0%, CER 98.1
- GPT-4.1: Acc@5 0.0%, CER 69.1
- Nanonets-OCR2-3B: Acc@5 0.0%, CER 95.0
- olmOCR-2: Acc@5 0.0%, CER 67.6
- RolmOCR: Acc@5 0.0%, CER 87.4