Welcome to the ComfyUI Serving Toolkit, a powerful tool for serving image generation workflows in Discord and other platforms (soon). This toolkit is designed to simplify the process of serving your ComfyUI workflow, making image generation bots easier than ever before. You can serve on discord, or on websockets.
If you need any help, Checkout the Discord!
- Allows both Images or videos (when in batch mode, such as animatediff - if you return more than one image it will create a video)
- Add arguments with default values, then allow your users to use them
- Serve from your own computer, workflow is not inserted into the images so your secrets are 100% safe
- Support for multiple serving options: Discord, Telegram, HTTP and WebSockets
Use ComfyUI Manager or put all the files inside custom_nodes and run:
..\..\..\python_embeded\python.exe -s -m pip install -r requirements.txt
Here a simple workflow that will get a !generate <prompt> and respond with an existing image, and !help will produce a help message
You can copy the workflow json: workflow.json
After setting up your workflow, In order for the serving to always be up, you need to allow auto queue, here is an image to help you do that:
DiscordServing
This node is an essencial part of the serving, queueing the prompt it will wait for a single message, process it and optionally return the image.
Note that in order for it to work for all messages you would have to mark Auto Queue (details above in the running section)
Inputs:
- discord_token - here is how you get one , make sure to enable message viewing intent
- command_name - the command used to generate, without the '!'. defaults to generate (so you would have to do !generate <your prompt> --your_argument1 <argument value>
Outputs:
- Serving Config - A basic reference for this serving, used by the other nodes of this toolkit to get arguments and return images.
WebsocketServing
This will connect to a websocket and wait for JSON of {_requestId, prompt, arguments} and will return a json of {_requestId, base64_img} You can see an example ws server over at examples/websocket.js to run it you would need to navigate to the examples folder and run these commands:
npm install
node websocket.js
Inputs:
- websocket_url - the url of the websocket you connect to, if you use the example it will be ws://localhost:8080
Outputs:
- Serving Config - A basic reference for this serving, used by the other nodes of this toolkit to get arguments and return images.
CommandPickerServing This node allows you to define specific behaviour for specific commands - it will only execute outputs that are connected to it when that specific command is given. Note, the command in HTTP Serving is the path after /, and the command in websocket is "command_name" json property
Inputs:
- serving_config - a config made by a serving node
- should_execute - an output of one of the output nodes, will only send that as part of the results if the command received is the one on this node
- command - the command to filter by
CommandPickerServing
AlwaysExecute For when you don't care about commands, it will always return an output regardless of command given. this will execute for ALL commands, no matter if the commands are registered or not.
Inputs:
- serving_config - a config made by a serving node
- should_execute - an output of one of the output nodes
ServingInputText
Allows you to grab a text arguments from the request
Discord example:
When a user types: !generate 4k epic realism portrait --negative drawing you could set the argument=negative and then recieve the value of "drawing" inside the output text.
Inputs:
- serving_config - a config made by a serving node
- argument - the argument name, the prompt itself will be inside the "prompt" argument. When using discord serving, you can access attachments url using 'attachment_url_0' (and attachment_url_1 etc). then you can use nodes like WAS Image Load to download these images
- default - the default value of this argument
Outputs: text - the value of the argument
ServingInputNumber
similar to ServingInputText, this one is for numbers. it is important to set the minimum, maximum and step to the right values in order to avoid errors (for example when trying a width that does isn't divisable by 16) Inputs that are not in ServingInputText:
- max_value - the maximum value of this argument
- min_value - the minimum value of this argument
- step - the steps of this value (setting this to 1 will ensure only whole numbers, 0.5 will allow jumps of half etc)
Serving Image/Video Output
Allows you to return an image/video back to the request Inputs:
- image - the generated image. note that if this is more than one image (for example in the case of batches or animatediff frames) it will return a video
- duration - in the case of a video, what is the time in miliseconds each frame should appear? if you have an FPS number you can use 1000/FPS to calculate the duration value
TelegramServing
This node allows you to serve your workflow via Telegram. It listens for commands and processes them, enabling interaction with your ComfyUI workflow through a Telegram bot.
Inputs:
- telegram_token - Your Telegram bot token obtained from the BotFather
- command_name - The command used to generate, without the '/'. Defaults to 'generate' (so you would have to do /generate --your_argument1 )
Outputs:
- Serving Config - A basic reference for this serving, used by the other nodes of this toolkit to get arguments and return images.
HTTPServing
This node enables serving your workflow via HTTP, allowing you to interact with your ComfyUI workflow through HTTP POST requests.
Inputs:
- port - The port number on which the HTTP server will run. Defaults to 8000.
Outputs:
- Serving Config - A basic reference for this serving, used by the other nodes of this toolkit to get arguments and return images.
ServingMultiImageOutput
Allows you to return multiple images back to the request. This is useful for batch processing or when generating animations.
Inputs:
- serving_config - A config made by a serving node
- images - The generated images
ServingInputImage
Allows you to input images from the serving platform (e.g., attachments in Discord, Telegram, or files sent via HTTP).
Inputs:
- serving_config - A config made by a serving node
- default_image - (Optional) A default image to use if no attachment is found
Outputs:
- image - The input image as a tensor
ServingInputImageAsLatent
Similar to ServingInputImage, but converts the input image directly into a latent representation for use in your workflow.
Inputs:
- serving_config - A config made by a serving node
- vae - The VAE model to use for encoding
- default_latent - (Optional) A default latent to use if no attachment is found
Outputs:
- latent - The input image encoded as a latent
