kadevin/ilab-gpt-conjure

面向 GPT-image-2 的 AI 图片生成 WebUI 工作台，支持 Codex Responses 与 OpenAI 兼容 API 接入，内置公用图库、多类型 Chip 快捷引用、提示词模板、多任务并发和本地队列管理。An AI image generation WebUI workbench for GPT-image-2 with Codex Responses and OpenAI-compatible API support, shared gallery references, multi-type quick chips, prompt templates, concurrent tasks, and local queue management.

What it solves

iLab CONJURE provides a centralized, local-first workbench for generating AI images across multiple models. It eliminates the need to switch between different AI interfaces by unifying GPT Image and Gemini into a single workspace, while adding professional management tools for prompts, reference images, and task history that are often missing from basic chat interfaces.

How it works

The project acts as a frontend and orchestration layer that connects to AI models via two primary methods: an OpenAI-compatible API mode (recommended for stability and teams) and an advanced local OAuth mode for personal Codex/ChatGPT workflows. It features a FastAPI-based WebUI and a CLI for automation. The system manages a local SQLite database for history and a local file system for storing generated images and reference libraries.

Who it’s for

It is designed for AI artists, designers, and power users who need a structured workflow for image generation, including the ability to manage large libraries of reference images, use complex prompt templates, and handle concurrent generation tasks.

Highlights

  • Multi-Model Support: Unified interface for GPT Image and Gemini, including text-to-image, image-to-image, and editing workflows.
  • Asset Management: A public image library for reusable reference assets and a prompt template system for structured generation.
  • Prompting Tools: Specialized "chips" for quick insertion of reference images (@), hex colors (#), and reusable prompt fragments (~).
  • Task Orchestration: Support for concurrent tasks, local queue management, and a paginated history library with search, filtering, and tagging.
  • Image Editor: Built-in editor supporting multi-layer composition, partial erasure, and locked-ratio transformations.
  • Deployment Options: Available as standard installers (DMG/EXE) or portable "one-click" packages for easy setup.