Note : PyTorch models in BaiduPan: https://pan.baidu.com/s/1r1DELT8BlgxeOP2RqREJEg code：6clx
PaddleOCR models in BaiduPan: https://pan.baidu.com/s/1getAprT2l_JqwhjwML0g9g code：lmv7
The downloadable models provided by PaddleOCR include inference model, trained model, pre-trained model and slim model. slim models are not supported here. The differences between the models are as follows:
| model type | model format | description |
|---|---|---|
| inference model | inference.pdmodel、inference.pdiparams | Used for reasoning based on Python prediction engine，detail |
| trained model, pre-trained model | *.pdparams、*.pdopt、*.states | The checkpoints model saved in the training process, which stores the parameters of the model, mostly used for model evaluation and continuous training. |
| model name | description | model size | download |
|---|---|---|---|
| ch_PP-OCRv4_det | [New] Original lightweight model, supporting Chinese, English, multilingual text detection | 4.70M | inference model / trained model |
| ch_PP-OCRv4_server_det | [New] Original general model, supporting Chinese, English, multilingual text detection | 110M | inference model / trained model |
| ch_PP-OCRv3_det | Original lightweight model, supporting Chinese, English, multilingual text detection | 3.8M | inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar) / trained model |
| ch_ppocr_mobile_v2.0_det | Original lightweight model, supporting Chinese, English, multilingual text detection | 3M | inference model / trained model |
| ch_ppocr_server_v2.0_det | General model, which is larger than the lightweight model, but achieved better performance | 47M | inference model / trained model |
| ch_PP-OCRv2_det | Original lightweight model, supporting Chinese, English, multilingual text detection | 3M | inference model / trained model |
| model name | description | model size | download |
|---|---|---|---|
| en_PP-OCRv3_det | [New] Original lightweight detection model, supporting English | 3.8M | inference model / trained model |
| model name | description | model size | download |
|---|---|---|---|
| ml_PP-OCRv3_det | [New] Original lightweight detection model, supporting English | 3.8M | inference model / trained model |
| model name | description | model size | download |
|---|---|---|---|
| ch_PP-OCRv4_rec | [New] Original lightweight model, supporting Chinese, English, multilingual text recognition | 10M | inference model / trained model |
| ch_PP-OCRv4_server_rec | [New] Original general model, supporting Chinese, English, multilingual text recognition | 88M | inference model / trained model |
| ch_PP-OCRv3_rec | Original lightweight model, supporting Chinese, English, multilingual text recognition | 12.4M | inference model / trained model |
| ch_ppocr_mobile_v2.0_rec | Original lightweight model, supporting Chinese, English and number recognition | 3.71M | inference model / trained model / pre-trained model |
| ch_ppocr_server_v2.0_rec | General model, supporting Chinese, English and number recognition | 94.8M | inference model / trained model / pre-trained model |
| ch_PP-OCRv2_rec | Original lightweight model, supporting Chinese, English, multilingual text detection | 8.5M | inference model / trained model |
Note: The trained model is finetuned on the pre-trained model with real data and synthsized vertical text data, which achieved better performance in real scene. The pre-trained model is directly trained on the full amount of real data and synthsized data, which is more suitable for finetune on your own dataset.
| model name | description | model size | download |
|---|---|---|---|
| en_PP-OCRv4_rec | [New] Original lightweight model, supporting english, English, multilingual text recognition | 9.7M | inference model / trained model |
| en_PP-OCRv3_rec | Original lightweight model, supporting english, English, multilingual text recognition | 9.6M | inference model / trained model |
| en_number_mobile_v2.0_rec | Original lightweight model, supporting English and number recognition(v2.0) | 2.56M | inference model / trained model |
| en_number_mobile_v2.0_rec | Original lightweight model, supporting English and number recognition(v2.1 the performance is optimized) | 2.6M | inference model / trained model |
| model name | description | model size | download |
|---|---|---|---|
| french_mobile_v2.0_rec | Lightweight model for French recognition | 2.65M | inference model / trained model |
| german_mobile_v2.0_rec | Lightweight model for German recognition | 2.65M | inference model / trained model |
| korean_mobile_v2.0_rec | Lightweight model for Korean recognition | 3.9M | inference model / trained model |
| japan_mobile_v2.0_rec | Lightweight model for Japanese recognition | 4.23M | inference model / trained model |
| it_mobile_v2.0_rec | Lightweight model for Italian recognition | 2.53M | inference model / trained model |
| es_mobile_v2.0_rec | Lightweight model for Spanish recognition | 2.53M | inference model / trained model |
| pt_mobile_v2.0_rec | Lightweight model for Portuguese recognition | 2.63M | inference model / trained model |
| ru_mobile_v2.0_rec | Lightweight model for Russia recognition | 2.63M | inference model / trained model |
| ar_mobile_v2.0_rec | Lightweight model for Arabic recognition | 2.53M | inference model / trained model |
| hi_mobile_v2.0_rec | Lightweight model for Hindi recognition | 2.63M | inference model / trained model |
| ch_tra_mobile_v2.0_rec | Lightweight model for chinese traditional recognition | 5.63M | inference model / trained model |
| ug_mobile_v2.0_rec | Lightweight model for Uyghur recognition | 2.63M | inference model / trained model |
| fa_mobile_v2.0_rec | Lightweight model for Persian recognition | 2.63M | inference model / trained model |
| ur_mobile_v2.0_rec | Lightweight model for Urdu recognition | 2.63M | inference model / trained model |
| rs_latin_mobile_v2.0_rec | Lightweight model for Serbian(latin) recognition | 2.53M | inference model / trained model |
| oc_mobile_v2.0_rec | Lightweight model for Occitan recognition | 2.53M | inference model / trained model |
| mr_mobile_v2.0_rec | Lightweight model for Marathi recognition | 2.63M | inference model / trained model |
| ne_mobile_v2.0_rec | Lightweight model for Nepali recognition | 2.63M | inference model / trained model |
| rs_cyrillic_mobile_v2.0_rec | Lightweight model for Serbian(cyrillic) recognition | 2.63M | inference model / trained model |
| bg_mobile_v2.0_rec | Lightweight model for Bulgarian recognition | 2.63M | inference model / trained model |
| uk_mobile_v2.0_rec | Lightweight model for Ukranian recognition | 2.63M | inference model / trained model |
| be_mobile_v2.0_rec | Lightweight model for Belarusian recognition | 2.63M | inference model / trained model |
| te_mobile_v2.0_rec | Lightweight model for Telugu recognition | 2.63M | inference model / trained model |
| kn_mobile_v2.0_rec | Lightweight model for Kannada recognition | 2.63M | inference model / trained model |
| ta_mobile_v2.0_rec | Lightweight model for Tamil recognition | 2.63M | inference model / trained model |
| model name | description | model size | download |
|---|---|---|---|
| french_mobile_v2.0_rec | Lightweight model for French recognition | 2.65M | inference model / trained model |
| german_mobile_v2.0_rec | Lightweight model for German recognition | 2.65M | inference model / trained model |
| korean_mobile_v2.0_rec | Lightweight model for Korean recognition | 3.9M | inference model / trained model |
| japan_mobile_v2.0_rec | Lightweight model for Japanese recognition | 4.23M | inference model / trained model |
| chinese_cht_mobile_v2.0_rec | Lightweight model for chinese cht recognition | 5.63M | inference model / trained model |
| te_mobile_v2.0_rec | Lightweight model for Telugu recognition | 2.63M | inference model / trained model |
| ka_mobile_v2.0_rec | Lightweight model for Kannada recognition | 2.63M | inference model / trained model |
| ta_mobile_v2.0_rec | Lightweight model for Tamil recognition | 2.63M | inference model / trained model |
| latin_mobile_v2.0_rec | Lightweight model for latin recognition | 2.6M | inference model / trained model |
| arabic_mobile_v2.0_rec | Lightweight model for arabic recognition | 2.6M | inference model / trained model |
| cyrillic_mobile_v2.0_rec | Lightweight model for cyrillic recognition | 2.6M | inference model / trained model |
| devanagari_mobile_v2.0_rec | Lightweight model for devanagari recognition | 2.6M | inference model / trained model |
For more supported languages, please refer to :
| model name | description | model size | download |
|---|---|---|---|
| ch_ppocr_mobile_v2.0_cls | Original model | 1.38M | inference model / trained model |