Model reference · open weights
bros-uncased is an open-weight embedding model from naver-clova-ocr. bros-base-uncased (BF16) weighs 438 MB; the smallest configuration that runs it is RTX 3060 12 GB.
For self-hosting, compare this build's requirements below with the GPU you plan to rent.
Summary of the naver-clova-ocr/bros-base-uncased model card, 2026-10-01
What it is
| Released by | naver-clova-ocr |
|---|---|
| Released | 2022-03-02 |
| VRAM | 438 MB for the weights |
What it runs on
| Card | Runs | Memory |
|---|---|---|
| RTX 3060 12 GB | fits | 11.6 GB |
| RTX 4060 Ti 16 GB | fits | 15.4 GB |
| RTX 3090 24 GB | fits | 23.4 GB |
| RTX 4090 24 GB | fits | 23.4 GB |
| RTX 5090 32 GB | fits | 31.0 GB |
| L40S 48 GB | fits | 44.0 GB |
| A100 80 GB | fits | 78.2 GB |
| H100 80 GB | fits | 78.1 GB |
| RTX PRO 6000 Blackwell 96 GB | fits | 93.8 GB |
| DGX Spark (GB10) 128 GB unified | fits | 107 GB |
| H200 141 GB | fits | 138 GB |
| B200 180 GB | fits | 176 GB |
From the model card
GitHub: https://github.com/clovaai/bros
BROS (BERT Relying On Spatiality) is a pre-trained language model focusing on text and layout for better key information extraction from documents. Given the OCR results of the document image, which are text and bounding box pairs, it can perform various key information extraction tasks, such as extracting an ordered item list from receipts. For more details, please refer to our paper:
BROS: A Pre-trained Language Model Focusing on Text and Layout for Better Key Information Extraction from Documents Teakgyu Hong, Donghyun Kim, Mingi Ji, Wonseok Hwang, Daehyun Nam, Sungrae Park AAAI 2022 - Main Technical Track
| name | # params | Hugging Face - Models |
|---|---|---|
| bros-base-uncased (this) | < 110M | naver-clova-ocr/bros-base-uncased |
| bros-large-uncased | < 340M | naver-clova-ocr/bros-large-uncased |
Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.