AI Guides

What Is iLab GPT Conjure? A Beginner‑Friendly Guide to kadevin/ilab-gpt-conjure

iLab GPT Conjure is a browser‑based WebUI for generating images with GPT‑image‑2, offering prompt templates, a reference‑image gallery, history management, CLI automation and OpenAI‑compatible API integration.

Published: Jun 14, 2026Updated: Jun 14, 2026Reading time: 6 minViews: 3

iLab GPT ConjureGPT-image-2WebUIPrompt TemplatesImage GalleryCLIOpenAI-compatible APIAI Image GenerationBeginner Guide

💡Key Takeaways

iLab GPT Conjure is a browser‑based WebUI for generating images with GPT‑image‑2, offering prompt templates, a reference‑image gallery, history management, CLI automation and OpenAI‑compatible API integration.

Repository: https://github.com/kadevin/ilab-gpt-conjure
Topic: GPT-image-2, AI image WebUI, prompt templates, gallery, image editing, OpenAI-compatible API
Audience: AI image creators, thumbnail/poster designers, people who need local image workflows, and developers who want image generation automation
Level: beginner-friendly, with minimal jargon
Image format: screenshot from the repository, no base64 embedding

1. What is iLab GPT Conjure in simple words?

iLab GPT Conjure is a WebUI for generating images with GPT-image-2.

Simple explanation:

TEXT

iLab GPT Conjure = a browser-based workbench for writing prompts, adding reference images, generating AI images, managing history, and downloading results.

If you use an image generation API directly, you usually need to write code or run commands. iLab GPT Conjure turns that workflow into a more visual interface.

Think of it as a local AI image studio:

Example

Write a prompt. Add reference images if needed. Choose image count, size, and quality. Start generation. Review results. Save or download the best images.

2. What problem does this repository solve?

AI image generation often becomes messy when you use it heavily.

Common problems:

Example

you reuse similar prompts many times reference images are scattered generated results are hard to organize you want to run several tasks at once you want to retry failed tasks you want to search old results you prefer a UI instead of code

iLab GPT Conjure is built to help with these problems.

It includes:

Example

reference image gallery prompt templates prompt snippets history page concurrent tasks CLI for automation portable packages for easier startup

3. What does “WebUI workbench” mean?

The README calls it an AI image generation WebUI workbench.

Simple meaning:

TEXT

WebUI = an interface that runs in the browser.
Workbench = a workspace with multiple helpful tools.

So iLab GPT Conjure is not just one prompt box. It is a workspace for managing the full image generation process.

Example:

Example

You have a fixed character reference. You save it in the gallery. You reuse it in multiple prompts. You save templates for posters, thumbnails, and product images. You generate many variations. You choose the best results. You check older work in history.

4. How can you use iLab GPT Conjure?

The repository supports two main usage modes.

Mode 1: WebUI

This is the best option for most users.

You open the local browser interface, write a prompt, add reference images, configure the image settings, and start generation.

Mode 2: CLI

CLI means command-line interface.

The README gives an example like:

Example

.venv/bin/python -m codex_image --prompt "A clean product photo of a ceramic mug" --out output/mug.png

This is useful for automation, batch generation, or scripts.

5. Two connection modes: API and local OAuth

This is the most important part to understand.

The README says the project supports:

Example

OpenAI-compatible API Codex / ChatGPT OAuth local mode

OpenAI-compatible API

This is the recommended mode for:

Example

stable integrations shared workstations team deployments public services production-like usage

You configure:

Example

Base URL API Key model name API mode concurrency

Simple meaning:

Example

This is the cleaner and more stable integration path.

Codex / ChatGPT OAuth local mode

This mode can reuse a local Codex / ChatGPT OAuth session to call an internal ChatGPT backend endpoint.

The README clearly warns that:

Example

this is not an officially recommended OpenAI API integration path the endpoint may change or stop working it may be subject to account, product, or usage restrictions it is intended only for personal local workflows

So for public writing, the safe message is:

Example

Use API mode for stability. Do not use local OAuth mode for production, team usage, or public services.

6. Main features

Text-to-image

You write a text prompt describing the image you want.

Example:

Example

A clean product photo of a ceramic mug on a white table, soft natural light.

Then the system sends the request to an image model.

Reference-image generation

You can add reference images.

Examples:

Example

character reference product photo logo style reference layout reference

This helps when you want consistent characters, products, or styles.

Image editing

The README says the project supports image editing workflows, so it is not only for creating images from scratch.

Concurrent tasks

The project supports concurrent tasks and local queue state.

Simple meaning:

Example

You can submit several jobs. Some jobs run now. Other jobs wait in the queue.

This is useful when generating many variations.

History page

The repository includes an independent /history page with pagination, search, filters, grid/list views, and lazy detail loading.

For real users, history matters because image generation often requires many attempts before choosing the best result.

Public gallery

The gallery is a local reusable reference library.

You can store:

Example

fixed characters products logos brand assets style references images you reuse often

Then you can insert them later without uploading again.

Prompt templates

Prompt templates are reusable long prompt structures.

Examples:

Example

YouTube thumbnail prompt product image prompt ad poster prompt consistent character prompt

Templates save time and make outputs more consistent.

Prompt chips

The prompt editor supports three types of chips:

Example

@ gallery chip # color chip ~ snippet chip

Simple meaning:

Example

@ inserts a gallery image. # inserts a color such as #FF6600. ~ inserts a reusable prompt snippet.

These make prompts more organized and reusable.

7. What are portable packages?

The README says the project provides portable packages.

Simple meaning:

Example

Download the package. Extract it. Run the start script. Open the local browser URL.

The README describes these packages as a ComfyUI-style unzip-and-run experience.

Portable packages include bundled CPython, dependencies, application source, license files, and a local data/ directory for settings, gallery files, inputs, outputs, task databases, and logs.

Important reminder:

Example

Personal data and generated images are stored locally in data/. Do not commit those files to Git.

8. Manual installation

According to the README:

BASH

git clone https://github.com/kadevin/ilab-gpt-conjure.git
cd ilab-gpt-conjure
python3 -m venv .venv
.venv/bin/python -m pip install -r requirements-webui.txt

Manual WebUI start:

Example

.venv/bin/python -m uvicorn codex_image.webui.app:app --host 127.0.0.1 --port 8787 --no-access-log

Then open:

Example

http://127.0.0.1:8787/

Requirements include:

TEXT

Python 3.11 or newer
requirements-webui.txt dependencies
frontend tooling if editing TypeScript or CSS

9. Who should use iLab GPT Conjure?

Frequent AI image creators

If you generate many images and need history, references, and templates, this WebUI can organize your workflow.

Thumbnail and poster designers

You can save prompt templates, brand colors, character references, and product references.

Product image creators

The gallery can store product references, and templates can keep a consistent visual style.

Character or visual concept creators

If you reuse the same character or style, the gallery and reference workflow are useful.

Developers who need automation

The CLI can be used in scripts or internal pipelines.

10. Who may not need this repository?

You may not need iLab GPT Conjure if:

Example

you only generate a few images occasionally you do not use GPT-image-2 or a compatible API you do not want to run a local app you do not need gallery, history, or templates you only want a simple online image generator you do not want to manage API keys or providers

For very simple needs, an existing hosted image tool may be enough.

11. Strengths

Example

Browser-based WebUI. CLI for automation. Reusable reference gallery. Prompt templates and snippets. History page for old results. Concurrent task handling. Portable packages for easier startup. OpenAI-compatible API mode. Chinese and English interface support.

12. Things to be careful about

Do not use local OAuth mode for production

The README says local OAuth mode is only for personal local workflows and is not an officially recommended OpenAI API integration path.

Do not commit sensitive data

Do not commit:

Example

OAuth files API keys input images generated outputs task metadata SQLite databases debug logs

AGPLv3 matters for network services

The project uses GNU AGPLv3. The README says that if you modify the software and make it available to users over a network, you must make the corresponding source code available under the same license.

macOS portable packages are unsigned

The README says macOS packages are unsigned portable zips, not signed or notarized apps. macOS may require right-click Open or quarantine removal.

The software license does not grant rights to model/API services

The README states that the license applies to the project code. It does not grant rights to the project name, logo, personal assets, API credentials, user prompts, input images, output images, or the model/API services used with it.

13. Simple comparison with ComfyUI

The README says portable packages are intended for a ComfyUI-style unzip-and-run experience.

But the tools are not the same.

Simple distinction:

Example

ComfyUI is often strong for node-based image workflows and local models. iLab GPT Conjure focuses on GPT-image-2 WebUI workflows, gallery, templates, history, and API-compatible image generation.

If you need deep node workflow control, ComfyUI may fit better. If you need a tidy GPT-image-2 workbench, iLab GPT Conjure is worth trying.

14. Example workflow for thumbnails

Example

Step 1: Save logo, character, or product references into the gallery. Step 2: Create a reusable thumbnail prompt template. Step 3: Use # color chips to keep brand colors. Step 4: Use @ gallery chips to insert reference images. Step 5: Generate several variations. Step 6: Choose the best result in the preview area. Step 7: Save and review results in history.

15. Example workflow for product images

Example

Step 1: Upload a product reference image. Step 2: Save it into the gallery. Step 3: Write a prompt for background, lighting, and camera angle. Step 4: Generate multiple versions. Step 5: Retry if part of the task fails. Step 6: Download the best image.

16. How beginners should read the repository

Do not read the whole repository from top to bottom.

Suggested path:

Example

Step 1: Read README.en.md or README.md. Step 2: Check Releases if you want a portable package. Step 3: Run the local WebUI. Step 4: Try API mode first. Step 5: Test gallery and prompt templates. Step 6: After that, explore the CLI. Step 7: If you want to modify code, read Development and License.

17. Conclusion

iLab GPT Conjure is a WebUI workbench for GPT-image-2. It is useful for people who want a more organized image generation workflow with reference images, gallery, prompt templates, history, concurrent tasks, and CLI automation.

Shortest explanation:

TEXT

iLab GPT Conjure = a local GPT-image-2 image studio with WebUI, gallery, prompt templates, history, and API-compatible mode.

Key reminders:

Example

Use OpenAI-compatible API mode for stable integration. Use local OAuth mode only for personal local workflows. Do not commit API keys, OAuth files, input/output images, databases, or logs. Read AGPLv3 carefully if you modify and offer the software over a network.

SEO title suggestions

What Is iLab GPT Conjure? Beginner-Friendly Guide to kadevin/ilab-gpt-conjure
Understanding iLab GPT Conjure: GPT-image-2 WebUI With Gallery and Prompt Templates
What Is iLab GPT Conjure Used For? WebUI, CLI, History, and Image Workflows
GPT-image-2 WebUI Workbench: iLab GPT Conjure Explained for Beginners

SEO meta description

A beginner-friendly guide to kadevin/ilab-gpt-conjure: what iLab GPT Conjure is, how it helps generate GPT-image-2 images, WebUI, CLI, gallery, prompt chips, templates, history, portable packages, API mode, local OAuth mode, strengths, limitations, and AGPLv3 license notes.

References

GitHub — kadevin/ilab-gpt-conjure: https://github.com/kadevin/ilab-gpt-conjure
README English — iLab GPT Conjure: https://github.com/kadevin/ilab-gpt-conjure/blob/main/README.en.md
README Chinese — iLab GPT Conjure: https://github.com/kadevin/ilab-gpt-conjure/blob/main/README.md
GitHub Releases — iLab GPT Conjure: https://github.com/kadevin/ilab-gpt-conjure/releases
Release v0.3.6: https://github.com/kadevin/ilab-gpt-conjure/releases/tag/v0.3.6

Written by PixelRouter Editorial Team

We publish deep, authoritative guides on AI infrastructure, API gateway security, cloud financial management, and system optimizations for developers.

FAQ

What is iLab GPT Conjure?

iLab GPT Conjure is a browser‑based WebUI workbench for generating images with GPT‑image‑2. It provides a visual interface for writing prompts, adding reference images, configuring image settings, reviewing results, and downloading the best outputs.

What connection modes does iLab GPT Conjure support?

The project supports two modes: an OpenAI‑compatible API mode for stable, production‑grade integrations, and a local Codex/ChatGPT OAuth mode intended only for personal, local workflows and not recommended for public or team use.

How can I install iLab GPT Conjure manually?

Clone the repository, create a Python 3.11+ virtual environment, install the web UI requirements, and start the server with uvicorn. Example steps: ``` git clone https://github.com/kadevin/ilab-gpt-conjure.git cd ilab-gpt-conjure python3 -m venv .venv .venv/bin/python -m pip install -r requirements-webui.txt .venv/bin/python -m uvicorn codex_image.webui.app:app --host 127.0.0.1 --port 8787 --no-access-log ``` Then open http://127.0.0.1:8787/ in a browser.

What are the main features of the iLab GPT Conjure WebUI?

Key features include text‑to‑image generation, reference‑image gallery, prompt templates and chips, image editing, concurrent task handling, a searchable history page, and portable packages for easy startup.

📂Related posts

AI Guides

AI Coding Agents Are Moving Beyond Code: The Visual Feedback Loop for iOS Development

An accessible analysis of visual feedback loops for iOS coding agents: editing Swift and SwiftUI, building apps, running iOS Simulator, observing the interface, hot reloading changes, attaching feedback to UI elements, and revising code.

👁 010 min

AI Guides

What Is blind-watermark? A Beginner-Friendly Guide to guofei9987/blind_watermark

A beginner-friendly guide to guofei9987/blind_watermark: what invisible watermarking is, how DWT–DCT–SVD works, Python and CLI usage, text/image watermark modes, crop and compression robustness, security limitations, privacy considerations, and the MIT License.

👁 06 min

AI Guides

What Is Orb.Farm? Browser-Based Aquatic Ecosystem Simulation

Learn what Orb.Farm is, how browser-based aquatic ecosystem simulation works, and what it teaches about algae, daphnia, fish, oxygen, CO₂, and ecological balance.

👁 18 min

← PixelRouter Blog