PaddleOCR-VL
PaddleOCR-VL is a vision-language model built around PaddleOCR that reads documents, forms, tables, charts, and screenshots. It combines strong OCR with reasoning over layout and content, then answers in text or structured JSON for multimodal RAG and automation.
- Type
- Multimodal
- Weights
- Closed
- Downloads
- 7.6k
- Likes
- 1.7k
- Pipeline
- image text to text
- License
- apache-2.0
Model hub
PaddlePaddle/PaddleOCR-VL
- Repo
- PaddlePaddle/PaddleOCR-VL
- Library
- PaddleOCR
- Access
- Public
Technical tags and languages
Papers
1
More from Baidu
12
-
ERNIE 3.0
Text
AI model by Baidu
-
ERNIE 4.5 Turbo
Text
ERNIE 4.5 Turbo is Baidu’s high-throughput, cost-optimized variant of ERNIE 4.5. It delivers strong reasoning…
-
684 DL
ERNIE 4.5 VL 28B A3B Thinking
Text
A multimodal MoE model that “looks, reads, and reasons” across images, video, and text. It adds tool use and…
-
2k DL
ERNIE 4.5-21B-A3B
Text
ERNIE 4.5-21B-A3B is Baidu’s efficient MoE variant of ERNIE 4.5—about 21B total parameters with ~3B active pe…
-
369 DL
ERNIE 4.5-300B-A47B
Text
ERNIE 4.5-300B-A47B is Baidu’s frontier Mixture-of-Experts ERNIE 4.5 variant—~300B total parameters with ~47B…
-
ERNIE 5
Text
ERNIE 5 is Baidu’s next-gen general model for reasoning, coding, and multimodal understanding. It supports lo…
-
ERNIE 5.1
Text
ERNIE-5.1 is Baidu’s new preview flagship language model, built for stronger general text capability, better…
-
ERNIE Bot
Text
AI model by Baidu
About Baidu
1
Baidu is a Chinese multinational technology company specializing in internet-related services, products, and artificial intelligence.