# Telegram — Download a message's media

Automatically download a message's media on web.telegram.org. Download the document, photo or video attached to a Telegram message, given its conversation peer id and message id. Returns the file as base64 plus the filename Telegram stores it under, and refuses a message that carries no media. Opening the conversation marks it read, and Telegram also saves the file to the browser's download folder. Photos come back as Telegram's re-encoded JPEG, not the original upload; documents come back byte-for-byte.

- Site: web.telegram.org
- Address: `reduck/web.telegram.org/download_media`
- Updated: 2026-10-23 (v2)
- Author: Reduck AI (reduck)

## Run it

Ask an AI agent connected to Reduck over MCP (`https://mcp.reduck.ai/mcp`) to run `reduck/web.telegram.org/download_media`, or run it from a terminal:

```bash
npx @reduck-ai/cli@latest run --script reduck/web.telegram.org/download_media
```

## Input

- `peerId` (string, required): The conversation's peer id, as returned by get_inbox (e.g. "8591421602" for Saved Messages; negative ids are channels and groups).
- `messageId` (string, required): The message carrying the media, as returned by get_conversation or send_message. Must be one of the recently loaded messages in that conversation — get_conversation only loads a recent window, so raise its count until the target is present.

## Output

- `kind` (string, required): Which bubble shape it was read from: "document" for a file bubble (a filename and size are shown in the chat), "media" for a photo or video bubble.
- `base64` (string, required): The file's bytes, base64-encoded — the exact inverse of send_message's fileBase64, so a file can be round-tripped between conversations. Documents come back byte-for-byte identical to what was sent.
- `peerId` (string, required): The peer id that was actually opened, read back from the address bar.
- `filename` (string, required): The name Telegram saves the file under, taken from its own download action rather than guessed. For a document this is the original filename; for a photo it is Telegram's generated name (e.g. "5793955541833946033.jpg").
- `messageId` (string, required): The message the media came from.
- `byteLength` (integer, required): Size of the returned bytes. For a document this matches the size shown on the bubble; for a photo it is the size of Telegram's re-encoded copy, which will not match the original upload.
- `verified_on_page` (boolean, required): True when the bytes were actually captured from the app's own download action, rather than the click being assumed to have worked.
- `mimeType` (string | null, optional): The MIME type Telegram attached to the blob it saved, or null if it did not set one.
- `chatTitle` (string | null, optional): The conversation title from the header — echo this to a human, since a peer id is not recognisable.

## FAQ

### What does "Telegram — Download a message's media" do?

Download the document, photo or video attached to a Telegram message, given its conversation peer id and message id. Returns the file as base64 plus the filename Telegram stores it under, and refuses a message that carries no media. Opening the conversation marks it read, and Telegram also saves the file to the browser's download folder. Photos come back as Telegram's re-encoded JPEG, not the original upload; documents come back byte-for-byte.

### How do I automatically download a message's media on web.telegram.org?

Ask an AI agent connected to Reduck to run reduck/web.telegram.org/download_media, or run it from a terminal with the Reduck CLI: npx @reduck-ai/cli@latest run --script reduck/web.telegram.org/download_media

### Is there a web.telegram.org API to download a message's media?

You do not need one. "Telegram — Download a message's media" drives the real web.telegram.org pages in a browser, so it works whether or not web.telegram.org offers an API for this.

### What information do I need to provide?

Required: peerId, messageId.

### What does it return?

It returns kind, base64, peerId, filename, mimeType, chatTitle, messageId, byteLength, verified_on_page.

### Do I need to be logged in to web.telegram.org?

Yes. It acts as you on web.telegram.org: on your own Chrome it reuses your session, and on a Reduck-hosted browser it loads the web.telegram.org cookies saved by the Reduck extension.

### Does it change anything on web.telegram.org, or only read data?

It only reads. It looks things up on web.telegram.org and changes nothing there.

### How do I run it?

Ask an AI agent connected to Reduck to run reduck/web.telegram.org/download_media, or run it from a terminal with the Reduck CLI: npx @reduck-ai/cli@latest run --script reduck/web.telegram.org/download_media

### Who maintains it?

It is part of Reduck's official curated catalogue.

Source: https://reduck.ai/explore/scripts/reduck/web.telegram.org/download_media