FAQ

Frequently asked questions

Install, run, branch — answered short.

Install

Can I run it through the Codex desktop app?

No — it runs through the Codex CLI. If the CLI is missing, ima2-gen prompts you to install it on first run, then opens a login window. Approve and you are good to go.

How do I install ima2-gen?

Open Terminal (macOS / Linux) or PowerShell (Windows). Install globally with npm install -g ima2-gen, then run ima2 serve. Open http://localhost:3333 in your browser.

npm install -g ima2-gen

ima2 serve

What is npm? Is there a one-click install?

npm is the Node.js package manager. If you do not have it, use the installer for your OS — it helps set up Node when needed, installs ima2-gen, and starts the server.

macOS

curl -fsSL https://lidge-jun.github.io/ima2-gen/install-mac.sh | bash

Windows PowerShell

irm https://lidge-jun.github.io/ima2-gen/install-windows.ps1 | iex

Does it work on Windows / PowerShell?

Yes. Windows 10+ with PowerShell 5.1 or PowerShell 7 are supported. The public installer script is served from GitHub Pages and prints every command before running it.

Usage

When should I use Node mode or Canvas Mode?

Use Node mode to compare several directions from a result you like. Use Canvas Mode when the frame is close but needs targeted cleanup, background work, annotation, or alpha/matte export.

How many images can I generate at once?

Up to eight in parallel from a single Classic run. Node mode runs each child node as its own request, so you can fan out as many branches as your quota allows.

Can I attach reference images?

Up to five references per prompt. Drag, drop, or paste them into the composer. Large files are compressed before upload, so you do not have to pre-process them.

How does a Style sheet work?

Save medium, composition, mood, palette, and negative cues into a sheet, then attach the sheet to any prompt. Style sheets persist across sessions, so you can reuse the same visual direction tomorrow.

What is multimode?

Multimode starts several Classic candidates from one prompt and shows slot-by-slot progress. You can stop a run, inspect partial results, and continue from the best image.

Can I import prompts from GitHub?

Yes. The prompt library can import local prompt packs, GitHub folders, curated sources, and GPT-image hint packs, then index them locally for search and ranking.

Tech

Do I need an OpenAI API key?

No for the default GPT OAuth path. ima2-gen can also use an OpenAI API key or bundled Grok/progrok when you explicitly choose those providers.

Where do my images live?

Locally. Generated images and session logs live in your user-data folder. Inspect, sync, version — your call. Image files are not uploaded to a separate remote store.

What data leaves my machine?

The generation request goes to the selected provider: OpenAI through GPT OAuth/API, or xAI through bundled progrok for Grok. The result file and session log stay on your disk.

Does it work offline?

Generation needs the selected provider network path, so an internet connection is required for that step. Browsing the local gallery, organizing sessions, and editing style sheets work offline.

Is it open source?

Yes. ima2-gen is MIT licensed. The code lives on GitHub.

github.com/lidge-jun/ima2-gen