dots.ocr-1.5
by rednote-hilab
Multilingual document parsing — 100+ languages, unified layout + recognition
rednote-hilab/dots.ocr-1.5mixpeek://image_extractor@v1/rednote_dots_ocr_15_v1Overview
dots.ocr-1.5 is a unified document parsing model from Xiaohongshu (RedNote) that combines layout detection and content recognition in a single model. It supports 100+ languages and handles academic papers, financial reports, tables, and multilingual content.
Task switching via prompt alone means no pipeline reconfiguration — the same model handles layout analysis, text extraction, and table parsing depending on the instruction.
Architecture
1.7B parameter model. Unified architecture that performs layout detection and OCR in a single forward pass. Prompt-based task switching for different document understanding modes.
Mixpeek SDK Integration
mixpeek.ingest.from_url(url="s3://documents/report.pdf",collection="documents",feature_extractors=[{"type": "ocr","model": "mixpeek://image_extractor@v1/rednote_dots_ocr_15_v1"}])
Capabilities
- Multilingual OCR (100+ languages)
- Layout detection
- Table extraction
- Academic paper parsing
- Financial document processing
Use Cases on Mixpeek
Performance
Common Pipeline Companions
Specification
Build a pipeline with dots.ocr-1.5
Add this model to a processing pipeline alongside other extractors. Combine with retrieval stages for end-to-end search.
Open Studio