ThinkAny RAG Search is an open-source retrieval-augmented generation search API from the ThinkAny team: it combines web search results with large language models so applications can answer questions with sourced, up-to-date information. The repository is released under the Apache-2.0 license, has about 1,200 stars on GitHub, and was last updated in 2024.
Building an AI search feature normally means stitching together a search provider, a scraping layer, ranking logic and an LLM call. This project packages that pipeline as a ready-to-run API: a query goes in, the service searches the web (via a Serper API key you supply), retrieves and organizes the relevant pages, and returns an answer grounded in those sources. It is the open code behind ThinkAny, an AI search engine product, and is aimed at developers who want Perplexity-style answer experiences inside their own applications without assembling every piece themselves. Setup is deliberately light: clone the repository, drop your keys into an environment file, and start the service. The codebase is small and readable, which also makes it a popular starting point for learning how production RAG search systems are structured.
The rag-search code is free and open source under the Apache-2.0 license; you run it yourself and pay only for your own search and LLM API usage. The hosted ThinkAny search product at thinkany.ai offers its own plans for users who prefer a managed service.
ThinkAny RAG Search is for developers building AI search, question-answering or research features into their products, and for engineers who want a compact reference implementation of a RAG search pipeline.