DeepFetch Semantic Search

Example Use Cases

• 1

  Time-sensitive facts and current event information with source-backed lookup.

• 2

  Extract and search text from PDFs such as reports, filings, and manuals.

• 3

  That require semantic reranking of web content to obtain evidence-rich snippets instead of raw links.

Description

DeepFetch is a semantic web search server compatible with MCP clients like Claude Desktop, Gemini CLI, and Codex CLI. It integrates Kagi discovery, Scrapfly extraction, local ONNX reranking, and PDF-aware retrieval to deliver evidence-backed search results rather than just URLs. The server supports PDF text extraction and offers a simple deployment path via Docker with minimal configuration. It aims to enhance agent workflows by providing rich, source-backed information efficiently.