TypeScript library to export Discord channels as transcripts (HTML, Markdown, TXT) — built for large volumes.
Existing solutions (like discord-html-transcripts) embed images as base64 directly inside the HTML, producing files hundreds of megabytes in size that are impossible to open. TranScriptor solves this with configurable media strategies and a streaming pipeline that never loads the entire channel into memory.
- 3 output formats — HTML, Markdown, TXT
- 4 media strategies —
reference|download|inline|none - Fine-grained control — toggle images, videos, files, reactions, embeds, stickers, and threads independently
- Native streaming — batch processing, constant memory usage regardless of channel size
- Optional compression — WebP via
sharp(optional peer dep) - ZIP export — automatically archive the
/assets/folder - JSON manifest — lists every exported asset with metadata
npm install transcriptor
# or
pnpm add transcriptorPeer dependencies:
npm install discord.js
# Optional — image compression
npm install sharpimport { createTranscript } from 'transcriptor';
// Minimal export — original Discord URLs, no downloading
const buffer = await createTranscript(channel, {
format: 'html',
output: 'buffer',
});
// Full export with downloaded assets
await createTranscript(channel, {
format: 'html',
output: 'file',
outputPath: './exports',
include: {
images: true,
videos: false,
files: true,
reactions: true,
embeds: true,
stickers: false,
threads: false,
},
media: {
strategy: 'download', // download to /assets/
downloadPath: './assets',
maxSizeMB: 5,
compress: true,
compressQuality: 80,
concurrency: 5,
},
});
// From already-fetched messages
import { generateFromMessages } from 'transcriptor';
const transcript = await generateFromMessages(messages, channel, {
format: 'markdown',
output: 'string',
});| Strategy | Description | File size | Self-contained |
|---|---|---|---|
reference |
Original Discord URLs (default) | Minimal | No (expire ~14d) |
download |
Downloads assets to /assets/ |
Medium | Yes |
inline |
Base64 embedded | Very large | Yes |
none |
Text placeholder [Image: ...] |
Minimal | Yes |
interface TranscriptOptions {
format: 'html' | 'markdown' | 'txt';
output: 'buffer' | 'string' | 'stream' | 'file';
outputPath?: string; // required when output === 'file'
include?: {
images?: boolean; // default: true
videos?: boolean; // default: true
files?: boolean; // default: true
reactions?: boolean; // default: true
embeds?: boolean; // default: true
stickers?: boolean; // default: true
threads?: boolean; // default: false
};
media?: {
strategy?: 'reference' | 'download' | 'inline' | 'none';
downloadPath?: string;
maxSizeMB?: number; // default: 8
compress?: boolean; // default: false (requires sharp)
compressQuality?: number; // default: 80
concurrency?: number; // default: 3
zip?: boolean; // default: false — zip the assets folder
};
limit?: number; // default: -1 (all messages)
chunkSize?: number; // default: 1000
filter?: (msg: Message) => boolean;
poweredBy?: boolean; // default: true
footerText?: string; // e.g. 'Exported {count} messages'
filename?: string;
}MIT