PaddleOCR 3.5 adds a Transformers inference backend, enabling developers to seamlessly use its OCR and document parsing models within the Hugging Face ecosystem, lowering integration barriers for building applications like RAG.
LlamaIndex releases ParseBench, the first document parsing benchmark designed for AI Agents, revealing that the traditional OCR standard of 'human-readable' is insufficient for agents' strict requirement of 'absolute correctness'.
LlamaIndex rewrote its lightweight PDF parser LiteParse in Rust, enabling cross-language and cross-platform (including browser) operation with up to 100x performance gains, providing critical infrastructure for real-time AI applications.
LlamaIndex releases ParseBench, the first OCR benchmark for AI agents, alongside tools tackling structural loss and security in document parsing, marking a paradigm shift from text extraction to contextual understanding.
LlamaIndex introduces ParseBench, the first OCR benchmark designed specifically for AI agents, alongside open-sourcing a local document parsing server and a secure sandboxed CLI agent, signaling a shift in document processing towards agent-native infrastructure.