# mem[v]

> The memory layer for humans and AI agents. Upload anything documents, videos, audio, images, chats, files from cloud storage and mem[v] turns it into a searchable, queryable memory graph you (and your agents) can talk to.

This page is served as raw Markdown to agents and rendered for humans. Fetch it any time at `https://memv.md/` (agents receive `text/markdown`; `https://memv.md/llms.txt` and `https://memv.md/index.md` always return raw Markdown).

## What mem[v] does

- **Ingest everything**: PDFs, Word/PowerPoint, HTML, images, audio, video, imports from AWS S3, Google Cloud Storage, Azure Blob, Google Drive, Dropbox, and Notion.
- **Understand it**: documents are parsed with OCR, videos and audio are analyzed with frontier video-language models, and everything is broken into temporally-aware segments.
- **Build a memory graph**: entities, relationships, and episodes are extracted into a knowledge graph with semantic search on top.
- **Use it anywhere**: chat with your memories, search them via API, or plug them into any AI agent over MCP.

## MCP server

mem[v] exposes your memory layer to AI agents through the Model Context Protocol.

- **Endpoint**: `https://mcp.memv.ai/mcp`
- **Transport**: Streamable HTTP

Example configuration (Claude Desktop / Claude Code / any MCP client):

```json
{
  "mcpServers": {
    "memv": {
      "url": "https://mcp.memv.ai/mcp"
    }
  }
}
```

Sign in at [app.memv.ai](https://app.memv.ai) to create an account and complete the MCP connection for your workspace.

## Surfaces

| URL | What it is |
|---|---|
| [memv.ai](https://memv.ai) | Main website |
| [app.memv.ai](https://app.memv.ai) | Developer dashboard : projects, spaces, uploads, memory graph, integrations |
| [chat.memv.ai](https://chat.memv.ai) | End-user portal : chat with your memories |
| [mcp.memv.ai](https://mcp.memv.ai) | MCP endpoint for agents |
| [memv.md](https://memv.md) | This page : agent-readable overview |

## Integrations

Connect once, import continuously: **AWS S3 · Google Cloud Storage · Azure Blob Storage · Google Drive · Dropbox · Notion**. Imported files flow through the same understanding pipeline as uploads (Notion pages are converted to documents automatically).

## For developers

- Projects and spaces organize memories; per-project settings control processing (including the video understanding model).
- Search API returns temporally-grounded facts extracted from your content.
- Activity, entity counts, and processing logs are visible in the dashboard.

## Status

Actively developed and in production. This file is the canonical machine-readable overview; it is updated as the platform evolves.

---

mem[v] · [memv.ai](https://memv.ai) · [app.memv.ai](https://app.memv.ai) · contact: [updates@info.memv.ai](mailto:updates@info.memv.ai)