All Models

PaddlePaddle PaddleOCR VL 1.5

AttachmentsOpen Weights

Multimodal model for analyzing text, images, documents, and rich media

Providers1
ReleasedJan 29, 2026
Input Modalitiestext, image
Output Modalitiestext

Available Providers (1)

ProviderModel IDInput CostOutput CostContextMax OutputDocs
SiliconFlow (China)PaddlePaddle/PaddleOCR-VL-1.5$0/MTok$0/MTok16.4K16.4K

Capabilities

Reasoning
Tool Calling
Attachments
Open Weights
Structured Output