June 2026 monthly performance highlights focused on delivering business value through robust OCR workflow modernization, cross-repo consistency, and maintainable code. Key initiatives reduced maintenance burden, improved reliability, and enabled scalable document processing across PDFs and images by aligning PaddleOCR integrations with official interfaces and API-driven patterns.

May 2026 — PaddleOCR (paddlepaddle/paddleocr): Delivered a cohesive, cross-language API experience and upgraded model capabilities to drive business value. Implemented PaddleOCR Multi-language API SDKs (Python, Go, TypeScript) that wrap the PaddleOCR async job API into simple blocking interfaces (submit → poll → fetch results) across OCR and document parsing tasks. Upgraded PaddleOCR-VL to model version 1.6 to improve accuracy and feature coverage. Hardened SDKs with robust error handling and a consistent error hierarchy (PaddleOCRAPIError), and reduced dependencies (Go/TS SDKs with zero extra dependencies). Introduced exponential backoff polling, URL/local file uploads, and resource-preserving defaults. CLI enhancements enable model-aware configuration and support for both OCR and document parsing tasks. Expanded OCROptions/DocParsingOptions to align with API parameters and ensured cross-language consistency in job definitions and defaults. These changes reduce integration time for downstream apps, improve reliability of batch OCR/doc parsing pipelines, and enable multi-language adoption of PaddleOCR in production workflows.