OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Search & Research AI

Minisearch

Minisearch is a self-hosted web-search application that combines a bundled SearXNG metasearch instance, local result reranking and an AI assistant that can run inference in the browser.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Minisearch — OpenViST project profile in Search & Research AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Minisearch when you need a self-hosted search interface that combines metasearch with cited question answering and can keep AI inference in the browser. Its documented workflow also supports browser default-search integration and Raycast access.

Who it is for

  • Researchers seeking cited answers from retrieved web results
  • Teams building private internal search or documentation interfaces
  • Self-hosters who want AI-assisted search without a separate inference server
  • Users who want browser-local search history, cached results and chats

Documented uses

  • Aggregate web results through SearXNG
  • Rerank search results locally before displaying them
  • Generate cited answers from retrieved sources
  • Ask follow-up questions using conversation memory
  • Run AI inference in the browser on WebGPU or CPU
Before you choose

Requirements & limits

The project provides a hosted Hugging Face demo and documents self-hosting through a published Docker image, Docker Compose, source build or a duplicated Hugging Face Space. No account or pricing requirement is documented in the supplied project material.

  • Browser-local inference can use WebGPU where available or CPU otherwise.
  • The bundled SearXNG instance supplies aggregated web results, while a cross-encoder reranks them locally.
  • The standard setup is a single Docker container and does not require a separate inference server.
  • Search history, cached results and chats are stored in the browser according to the project description.
  • Remote inference remains available through OpenAI-compatible APIs or AI Horde when local browser inference is not selected.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up Minisearch with the published Docker image

Start a self-hosted Minisearch instance using the documented single-container setup.

Before you start

  • Docker
  • A machine able to run the published ghcr.io/felladrin/minisearch image

Steps

  1. Run the published Minisearch Docker image with port 7860 mapped to port 7860.

  2. Open http://localhost:7860 in a browser.

  3. Start searching through the web interface.

Expected outcome

A locally reachable Minisearch instance is available for web searching and its browser-based AI workflow.

Guide limitations

  • This guide covers the documented container startup and opening the interface, not an independent execution test.
  • The evidence does not specify hardware requirements for browser model performance.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
4 October 2026
Original website title
felladrin/MiniSearch · Official homepage title · retained in its source language
How we review sourcesSuggest a correction