Retrieval-Augmented Generation (RAG) Workflows in R with Local and Web Search R package RAGFlowChainR version 0.1.7 Enables Retrieval-Augmented Generation (RAG) workflows in R by combining local vector search using 'DuckDB' with optional web search via the 'Tavily' API. Supports 'OpenAI'- and 'Ollama'-compatible embedding models, full-text and 'HNSW' (Hierarchical Navigable Small World) indexing, and modular large language model (LLM) invocation. Designed for advanced question-answering, chat-based applications, and production-ready AI pipelines. This package is the R equivalent of the 'python' package 'RAGFlowChain' available at <https://pypi.org/project/RAGFlowChain/>. Software Depends: R (>= 4.1.0) Imports: DBI, duckdb (>= 0.10.0), httr, dplyr, pdftools, officer, rvest, xml2, curl, Suggests: testthat (>= 3.0.0), jsonlite, stringi, magrittr, roxygen2, knitr, rmarkdown, VectrixDB Kwadwo Daddy Nyame Owusu Boakye <kwadwo.owusuboakye@outlook.com> Comprehensive R Archive Network (CRAN) Kwadwo Daddy Nyame Owusu Boakye [aut, cre] MIT + file LICENSE (https://CRAN.R-project.org/package=RAGFlowChainR/LICENSE) 2026-02-22 application/tgz https://CRAN.R-project.org/package=RAGFlowChainR doi:10.32614/CRAN.package.RAGFlowChainR