---
title: "MnemoVi — Local Video Knowledge Base"
description: "Index what was said and what was on screen across your videos, search it, and jump to the moment. Runs locally; nothing is uploaded."
canonical: https://giglabo.com/products/mnemoria/applications/mnemovi-local-video-knowledge-base
locale: en
---

# MnemoVi — Local Video Knowledge Base

> Markdown twin of https://giglabo.com/products/mnemoria/applications/mnemovi-local-video-knowledge-base
> Fetch this instead of the HTML page: same content, a fraction of the bytes.
> Site structure and the full page list for agents: https://giglabo.com/llms.txt

Index what was said and what was on screen across your videos, search it, and jump to the moment. Runs locally; nothing is uploaded.

- Type: page

## What MnemoVi is

MnemoVi is a desktop application that turns your videos into a searchable, local knowledge base. It transcribes every word and describes every frame, so you can search both what was **said** and what was **on screen**, then jump to the exact second. Everything — speech recognition, frame analysis, search — runs on your own machine. No cloud, no uploads, no subscription.

Example queries it answers: "where did they explain the K8s OOM fix", "the slide with the sharding diagram", "the recipe where she salts the water twice".

## Key capabilities

- **Said / On-screen / Both** — every result is labeled by source, with an exact timecode.
- **Reads the screen like you do** — slides, diagrams, product labels, even terminal output. If it appeared in the frame, it's searchable.
- **Your AI agent can use it** — MnemoVi runs a local MCP server, so Claude Desktop, Claude Code, Codex, Gemini CLI or any MCP-capable agent can search and quote your video library, grounded in what was actually said and shown. See the [MnemoVi MCP server page](https://giglabo.com/products/mnemoria/applications/mnemovi-local-video-knowledge-base/mcp).
- **Nothing leaves your device** — speech, vision and search run locally. Verify it with a firewall.

## How it works

1. **Add** — drop in files, point MnemoVi at a folder, or paste a link from YouTube, TikTok, Instagram or X (fetched through your own browser session, straight to your disk).
2. **It watches and listens** — transcription and frame description run in the background, on your hardware. Search what's ready while the rest is processing.
3. **Ask and jump** — type what you remember in plain words, click a result, land on the exact second.

## Platforms and privacy

- Native desktop apps for macOS (Apple Silicon), Windows and Linux; uses the GPU to speed up indexing when available, works on CPU alone.
- Works fully offline. The license does not phone home; the index is an open on-disk format. If the vendor disappears, MnemoVi keeps working.
- The only things that ever leave the machine are user-initiated: the AI model download on first launch and the clips you paste links for.
- Transcription supports dozens of languages including English, German and Russian; search works in the language of your content.

## Pricing

- **Free for your first 5 videos** — every feature, no time limit. When you add video #6, MnemoVi asks for a license; your knowledge base and index carry over.
- **Pay once, own it** — lifetime license, all features included, no tiers, all future updates. Solo (1 license, regular €200) and Team (3 licenses, regular €400); time-limited launch discounts may apply on the page.
- No subscription, no credits, 14-day refund, price never grows with your library.

## Where to go next

- [Mnemoria product line overview](https://giglabo.com/products/mnemoria)
- [Connect MnemoVi to an AI agent (MCP)](https://giglabo.com/products/mnemoria/applications/mnemovi-local-video-knowledge-base/mcp)
- Use-case pages live under `/products/mnemoria/applications/mnemovi-local-video-knowledge-base/use-cases/…` — see the site map for the full list.

## Related

- HTML version of this page: https://giglabo.com/products/mnemoria/applications/mnemovi-local-video-knowledge-base
- Site map for agents: https://giglabo.com/llms.txt