midjourney-discord-api

midjourney-discord-api

Deno doc

midjourney-discord-api is a library designed to connect to a Discord channel and send messages to be processed by the Midjourney bot. It utilizes the same requests as the Discord web client, allowing seamless communication with the bot. To configure the library, extract an authenticated request sent to the Midjourney bot using your web development tools.

Features

feature	Status	feature	Status
/blend	✅	Reroll	✅
/describe	✅	/relax	✅
/fast	✅	/settings	✅
/imagine	✅	Variations	✅
Upscale	✅	concurent calls	✅
/private	❌ N/A	/public	❌ N/A
/show	❌ N/A	/stealth	❌ N/A

Installation

NodeJS ESM or CJS

npm install midjourney-discord-api

Deno

deno add @u4/midjourney
import Midjourney from "@u4/midjourney";

NodeJS (ESM) + JSR

npx jsr add @u4/midjourney
import Midjourney from "@u4/midjourney";

ESM nodeJS

import Midjourney from "midjourney-discord-api";
// or
// import Midjourney from "@u4/midjourney";

const cli = new Midjourney("interaction.txt");
const msgs = await cli.getMessages();
console.log(msgs.length + " messages visibles"); // by default get 50 messages

Deno

// deno add @u4/midjourney
import Midjourney from "@u4/midjourney";
// using x
// import Midjourney from "midjourney-discord-api";

Token / ids extraction.

• Open your web browser to https://discord.com/channels/@me
• Click in the discord channel you want to work into like https://discord.com/channels/662267976984297473/933565701162168371
• Open the developent bar (on chrome: F12 on windows Command ⌘ + Option ⌥ + I on mac)
• Go to network

• Send a request to discordBot, like /settings
• Left click on the https://discord.com/api/v9/interactions request or https://discord.com/api/v9/science.
• Click Copy
• Click Copy as fetch

• Save this request in a file, that you will provide to the Midjourney constructor. (for my Test I name this file interaction.txt)

Midjourney client initialization

the Midjourney client can be initilized in 4 different ways:

new Midjourney(filename) fetch command

in this case the provided file should contains a valid NodeJS fetch command as described in the previous section.

filename can also be replace by it's content.

new Midjourney(filename) environement file

in this case the provided file should contains somthink like:

SALAI_TOKEN = your 72 char long authorization token
SERVER_ID = the serverId
CHANNEL_ID = the channelId

SERVER_ID and CHANNEL_ID can be omit, if you provide those data using Midjourney.setDiscordChannelUrl("https://discord.com/channels/1234567890/1234567890")

environement variable

in this case SALAI_TOKEN should contains you 72 char long authorization token. SERVER_ID and CHANNEL_ID should contains the serverId and channelId

SERVER_ID and CHANNEL_ID can be omit, if you provide those data using Midjourney.setDiscordChannelUrl("https://discord.com/channels/1234567890/1234567890")

new Midjourney(SALAI_TOKEN) directly provide SALAI_TOKEN

in this case a 72 char long SALAI_TOKEN token should be pass to the Midjourney constructor like: new Midjourney(SALAI_TOKEN)

then you will have to provide serverId and channelId using Midjourney.setDiscordChannelUrl("https://discord.com/channels/1234567890/1234567890")

Usage

Here are some examples of how to use the Midjourney class:

Describe URL

import Midjourney from "midjourney-discord-api";
// or
// import Midjourney from "@u4/midjourney";

const client = new Midjourney("interaction.txt");
await client.connectWs(); // Used Websocket to boost detection. (experiental)
const prompts: string[] = await client.describeUrl(
  "https://cdn.midjourney.com/95e2c8fd-255c-4982-9065-83051143570c/0_0_640_N.webp",
  // add optional progress function (percent) => void
);
console.log("reversed prompt: ", prompts);

Imagine

import Midjourney from "midjourney-discord-api";
// or
// import Midjourney from "@u4/midjourney";

const client = new Midjourney("interaction.txt");
await client.connectWs(); // Used Websocket to boost detection. (experiental)
const msg = await client.imagine(
  "A photo of an astronaut riding a horse",
  // add optional progress function (percent) => void
);
console.log("you find your result here: ", msg.attachments[0].url);

Upscale

// Upscale the first none upscaled images in chat, searching from the newest to the oldest images
import Midjourney from "midjourney-discord-api";
// or
// import Midjourney from "@u4/midjourney";

// Variant the last image available in chat
const client = new Midjourney("interaction.txt");
await client.connectWs(); // Used Websocket to boost detection. (experiental)
const msgs = await client.getMessages();
main:
for (const msg of msgs) {
  if (!msg.canVariant()) {
    continue;
  }
  for (let i = 1; i <= 4; i++) {
    const v = msg.canVariant(i)
    if (v) {
      console.log(`Variant image ${v.custom_id} from ${msg.id}: ${msg.prompt?.prompt}`);
      const result = await msg.variant(i);
      await result.download(0, "images");
      break main;
    }
  }
}

Variant

// Variant the last image available in chat
const client = new Midjourney("interaction.txt");
await client.connectWs(); // Used Websocket to boost detection. (experiental)
const msgs = await client.getMessages();
main:
for (const msg of msgs) {
  if (!msg.canVariant()) {
    continue;
  }
  for (let i = 1; i <= 4; i++) {
    const v = msg.canVariant(i)
    if (v) {
      console.log(`Variant image ${v.custom_id} from ${msg.id}: ${msg.prompt?.prompt}`);
      const result = await msg.variant(i);
      await result.download(0, "images");
      break main;
    }
  }
}

Reroll

import Midjourney from "midjourney-discord-api";
// or
// import Midjourney from "@u4/midjourney";

const client = new Midjourney("interaction.txt");
await client.connectWs(); // Used Websocket to boost detection. (experiental)
const msg = await client.imagine(
  "A photo of an astronaut riding a horse",
  // add optional progress function (percent) => void
);
if (msg.canReroll()) {
  const result = msg.reroll();
  console.log(`upscale V2 Ready from`, result.attachments[0].url);
}

More Samples

check all the samples here

name	function
ImagineSet.ts	Call concurent /imagine function at the same time
blendImages.ts	blend 2 random images
describeRegen.ts	descript an image then imagine the returned prompts
listLastMsgs.ts	list last message
parseOne.ts	debug one message
progressLogger.ts	a progress logger provider
sampleUrls.ts	samples Image url used in test script
saveAll.ts	download all image from a channel
upscaleLast.ts	upscall the last non upscalled image
upscaleLasthoursAgo.ts	upscall an image at a specific time
variantLast.ts	generate variante for a the last generated image

Contributing

We welcome contributions to the midjourney-discord-api project. Please feel free to submit issues, feature requests, and pull requests to improve the project.

reference

• Discord Snowflake to Timestamp Converter

License

This project is licensed under the MIT License.