UnSearch – Open-source search API aggregating 70+ engines for AI agents

Summary: UnSearch is an open-source search API designed for AI agents, aggregating results from over 70 search engines and delivering clean, structured data for large language models. It offers a low-cost alternative to Tavily with zero data retention and options for self-hosting or cloud use.

What it does

UnSearch queries multiple search engines simultaneously and returns structured data optimized for LLM consumption. It supports zero-retention mode and can be self-hosted or accessed via the cloud.

Who it's for

It targets developers building AI agents, retrieval-augmented generation (RAG) pipelines, and LLM applications requiring real-time web search data.

Why it matters

It addresses outdated LLM training data by providing affordable, real-time web search with enhanced privacy and flexibility compared to existing solutions.