ComfyUI Quality of Life Suite

Thank you to all the valuable contributors. Kindly submit any pull requests to the development branch instead of the main branch. Your efforts are greatly appreciated.

A feature-rich ComfyUI extension providing 50+ quality-of-life nodes including OpenAI (ChatGPT & DALL-E 2) integration, text processing, image/latent scaling, math utilities, and more.

---

Table of Contents

• Features
• Installation
• Configuration
• Node Reference
  • OpenAI Suite
  • String Suite
  • Latent Tools
  • Image Tools
  • Number / Math Tools
  • Utility Nodes
  • Debug Nodes
• Example Workflows
• Fonts
• Auto-Update
• Contact

---

Features

• 🤖 OpenAI Integration — ChatGPT (GPT-3.5 / GPT-4) prompt generation and DALL-E 2 image creation
• 📝 String Suite — Concatenate, trim, replace, split, and save text; convert text to an image
• 🎲 NSP (Noodle Soup Prompts) — Random prompt term generation from categorised terminology
• 🖼️ Image Tools — Scale images using multiplication factors instead of fixed pixel dimensions
• 🔲 Latent Tools — Upscale latents by factor and select individual latents from a batch
• 🔢 Number / Math Tools — Evaluate equations, convert between types, read image/latent dimensions
• 🛠️ Utility Nodes — Thin wrappers for text, integer, float, and seed inputs
• 🐛 Debug Nodes — Write intermediate values to the console, optionally passing them through

All nodes appear in the ComfyUI menu under the O/ top-level category.

---

Installation

Option A — Manual (zip)

1. Download the repository as a zip file.
2. Extract the folder into <ComfyUI root>/custom_nodes/.
3. The resulting path should look like custom_nodes/ComfyUI-QualityOfLifeSuit_Omar92/.
4. Restart ComfyUI (a browser reload is not sufficient).

Option B — Git clone

cd <ComfyUI root>/custom_nodes
git clone https://github.com/omar92/ComfyUI-QualityOfLifeSuit_Omar92.git

Then restart ComfyUI.

After installation you will find all nodes grouped under O/… in the node search / right-click menu.

Tip: See how directory should look like.png at the repository root for the expected folder structure.

---

Configuration

On the first run ComfyUI will automatically create a config.json file inside the extension folder:

{
  "autoUpdate": true,
  "branch": "main",
  "openAI_API_Key": "sk-#################################"
}

Key	Default	Description
autoUpdate	true	Pull the latest changes from GitHub every time ComfyUI starts. Set to false to disable.
branch	"main"	Git branch to pull updates from. Use "dev" for the latest development changes.
openAI_API_Key	placeholder	Your OpenAI API key. Required for all ChatGPT and DALL-E nodes. Get yours at platform.openai.com/api-keys.

Security note: The API key is stored in config.json and is never embedded in generated images.

---

Node Reference

OpenAI Suite

Located under O/OpenAI and O/OpenAI/Advanced/….

Simple ChatGPT

ChatGPT Simple _O — O/OpenAI

The quickest way to use ChatGPT. Provide a short prompt and the node returns a fully formed Stable Diffusion prompt or a detailed image description.

Input	Type	Description
prompt	STRING	Your input text / subject
model	dropdown	OpenAI model to use (e.g. gpt-3.5-turbo)
behaviour	dropdown	tags — returns SD-style comma-separated tags; description — returns a natural language description
seed	INT	Controls randomness; same seed ≈ same output

Output	Type
result	STRING

ChatGPT compact _O — O/OpenAI

Mid-complexity ChatGPT node with a customisable system initialisation message.

Input	Type	Description
prompt	STRING	User message
initMsg	STRING	System message that shapes the AI's behaviour
model	dropdown	OpenAI model
seed	INT	Randomness seed

---

Advanced OpenAI

load_openAI _O — O/OpenAI/Advanced

Loads the OpenAI client. Required by all Advanced nodes.

Input	Type	Description
base_url	STRING	API endpoint (defaults to OpenAI; change for compatible providers)
api_key	STRING	Your OpenAI API key

Output	Type
client	CLIENT

Chat_Message _O — O/OpenAI/Advanced/ChatGPT

Creates a single chat message.

Input	Type	Description
role	dropdown	user, assistant, or system
content	STRING	Message text

Output	Type
messages	OPENAI_CHAT_MESSAGES

combine_chat_messages _O — O/OpenAI/Advanced/ChatGPT

Merges two message lists into one, allowing you to chain multi-turn conversations.

Chat completion _O — O/OpenAI/Advanced/ChatGPT

Sends a message list to the API and returns the model's reply.

Input	Type	Description
client	CLIENT	From load_openAI _O
model	dropdown	OpenAI model
messages	OPENAI_CHAT_MESSAGES	Built with Chat_Message / combine_chat_messages
seed	INT	Randomness seed

Output	Type
text	STRING
completion	OPENAI_CHAT_COMPLETION

---

DALL-E 2

create image _O — O/OpenAI/Advanced/Image

Generates images from a text prompt using DALL-E 2.

Input	Type	Description
client	CLIENT	From load_openAI _O
prompt	STRING	Image description
number	INT (1–10)	How many images to generate
size	dropdown	256x256, 512x512, or 1024x1024

Output	Type
IMAGE	IMAGE
MASK	MASK

variation_image _O — O/OpenAI/Advanced/Image

Creates variations of an existing image using DALL-E 2.

Input	Type	Description
client	CLIENT	From load_openAI _O
image	IMAGE	Source image
number	INT (1–10)	Number of variations
size	dropdown	Output resolution

---

String Suite

Located under O/text and O/text/operations.

Text _O — O/utils Simple pass-through text input node.

Concat Text _O — O/text/operations Concatenates up to 13 strings with a configurable separator (default: , ).

Trim Text _O — O/text/operations Removes leading and trailing whitespace from a string.

Replace Text _O — O/text/operations Replaces all occurrences of a substring with another string.

Input	Description
text	Input string
old	Substring to find
new	Replacement string

QOL Split String — O/text/operations Splits a string by a delimiter and exposes up to 13 individual outputs.

Input	Description
text	Input string
delimiter	Character(s) to split on

saveTextToFile _O — O/text Saves text to a file inside the ComfyUI /output folder. A timestamp is appended to the filename to avoid overwriting.

Input	Description
text	Content to save
filename	Base filename
append	If true, appends to an existing file instead of creating a new one

---

Text-to-Image

Text2Image _O — O/text

Renders text onto a canvas and outputs it as an IMAGE. Useful with ControlNet to place styled text into a generated image.

Input	Type	Description
text	STRING	Text to render
font	dropdown	Font selected from the fonts/ directory
size	INT (0–255)	Font size in points
font_R/G/B/A	INT (0–255)	Font colour (RGBA)
background_R/G/B/A	INT (0–255)	Background colour (RGBA)
width / height	INT	Canvas size in pixels
expand	bool	Auto-expand canvas to fit the text
x / y	INT	Text position on canvas

Custom fonts: Add any .ttf, .otf, or .ttc file to the fonts/ folder and restart ComfyUI — it will appear in the font dropdown automatically.

---

NSP (Noodle Soup Prompts)

Located under O/text/NSP.

NSP is a community-maintained collection of categorised prompt terms. The extension downloads the list automatically on first use.

RandomNSP _O Returns a random term from the chosen category.

Input	Description
terminology	Category (e.g. artist, style, subject, …)
seed	Randomness seed

ConcatRandomNSP_O Appends a random NSP term to an existing string — handy as an in-route modifier.

Input	Description
text	Existing string
terminology	NSP category
separator	String inserted between text and the NSP term (default , )
seed	Randomness seed

---

Latent Tools

Located under O/latent.

LatentUpscaleFactor _O Upscales a latent tensor using separate width and height multiplication factors.

Input	Description
samples	Input LATENT
upscale_method	Interpolation method
WidthFactor	Multiplier for width (e.g. 2 doubles the width)
HeightFactor	Multiplier for height
crop	Whether to crop after upscaling

Example: 512×512 latent with WidthFactor=2, HeightFactor=2 → 1024×1024. Use factors < 1 to downscale (e.g. 0.5 → 256×256).

LatentUpscaleFactorSimple _O Same as above but uses a single factor applied to both dimensions.

selectLatentFromBatch _O Picks a single latent from a batch by index. Useful when you generate multiple images and want to apply further processing to just one of them.

Input	Description
samples	Batch LATENT
index	Zero-based index of the latent to select

---

Image Tools

Located under O/image.

ImageScaleFactor _O Scales an IMAGE tensor using separate width and height multiplication factors.

Input	Description
image	Input IMAGE
upscale_method	Interpolation method
WidthFactor	Width multiplier
HeightFactor	Height multiplier
MulOf46	When enabled, rounds output dimensions to the nearest multiple of 8 — required by most Stable Diffusion models (enabled by default)
crop	Whether to crop after scaling

ImageScaleFactorSimple _O Same as above but with a single uniform scale factor.

GetImage_(Width&Height) _O — O/numbers Reads the pixel dimensions of an IMAGE and outputs them as separate INT values.

---

Number / Math Tools

Located under O/numbers.

Equation1param _O Evaluates a Python math expression with one variable x.

Input	Description
x	FLOAT value for x
equation	Expression string, e.g. x * 2 + 1

Outputs: FLOAT, INT

Equation2params _O Evaluates two expressions with variables x and y.

Input	Description
x	First FLOAT
y	Second FLOAT
equation	First expression, e.g. x + y
equation_2	Second expression, e.g. x * y

Outputs: FLOAT, INT, FLOAT, INT (one pair per equation)

floatToInt _O — Convert FLOAT → INT

intToFloat _O — Convert INT → FLOAT

floatToText _O — Convert FLOAT → STRING

seed2String _O — Convert a SEED value → STRING

GetLatent_(Width&Height) _O — Read width and height from a LATENT tensor as INTs

---

Utility Nodes

Located under O/utils.

Node	Output	Description
Text _O	STRING	Text input widget
seed _O	INT	Seed input widget
int _O	INT	Integer input widget
float _O	FLOAT	Float input widget
Note _O	(none)	Read-only sticky note for annotating workflows

---

Debug Nodes

Located under O/debug/….

Node	Description
Debug Text _O	Prints [prefix] text to the console. No output — terminal only.
Debug Text route _O	Same as above but also passes the string through, so it can be inserted mid-route.
Debug OpenAI Chat Messages _O	Prints the full message list to the console.
Debug OpenAI Chat Completion _O	Prints the full completion object to the console.

---

Example Workflows

The Workflows/ directory contains ready-to-use workflow files and screenshots:

File	Description
All nodes Workdflow .json	Master workflow demonstrating every node
ChatGPT.png	Simple ChatGPT-to-prompt workflow
ChatGPT_Advanced.png	Advanced multi-message ChatGPT workflow
string_o.png	String manipulation node examples

To load a workflow from an image, drag the PNG file onto the ComfyUI canvas.

Simple ChatGPT Workflow

1. Add a ChatGPT Simple _O node.
2. Type your subject in the prompt field (e.g. "a futuristic cityscape at night").
3. Choose tags behaviour to get Stable Diffusion prompt tags, or description for a natural language description.
4. Connect the output STRING to your KSampler positive conditioning.

Advanced ChatGPT Workflow

1. Add load_openAI _O and enter your API key (or rely on config.json).
2. Create a Chat_Message _O with role = system to set the AI's behaviour.
3. Create another Chat_Message _O with role = user for your actual prompt.
4. Use combine_chat_messages _O to merge both messages.
5. Feed the result into Chat completion _O to get the AI's reply as a STRING.

---

Fonts

The fonts/ folder ships with several built-in fonts:

File	Font
Alkatra.ttf	Alkatra
CALIBRI.TTF	Calibri
COMIC.TTF	Comic Sans
COMICI.TTF	Comic Sans Italic
COMICZ.TTF	Comic Sans Bold

To add a custom font, copy any .ttf, .otf, or .ttc file into the fonts/ directory and restart ComfyUI. The font will appear automatically in the dropdown of the Text2Image _O node.

---

Auto-Update

The extension automatically checks for updates every time ComfyUI starts (when autoUpdate is true in config.json). It uses pygit2 to pull the latest commits from the configured branch.

To disable auto-update, open config.json and set:

"autoUpdate": false

To switch to the development branch for the latest (potentially unstable) features:

"branch": "dev"

---

Thanks for reading, and I hope my tools will help you!

Contact

• Discord: Omar92#3374
• GitHub: omar92
• CivitAI: omar92