luckyabsoluter/codex-imagegen-free-reference — explained in plain English
Analysis updated 2026-05-18
Choose a specific local reference image instead of letting Codex pick one automatically for image generation.
Generate or edit images using Codex-auth models like gpt-image-2 that the standard OpenAI API path cannot reach.
Pick between the Responses hosted image route and the Image API route for a given generation task.
| luckyabsoluter/codex-imagegen-free-reference | 0marildo/imago | 0xdfi/glm-5.2-1m-4x-dgx-spark | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | Python | Python | Python |
| Setup difficulty | easy | easy | hard |
| Complexity | 2/5 | 2/5 | 5/5 |
| Audience | developer | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires an existing Codex installation and account, no separate API key mentioned.
Codex ImageGen Free Reference is an add-on for a built-in image generation feature already included in Codex, an AI coding tool. The built-in feature creates or edits images automatically but manages reference images on its own without letting you pick them. This project changes that by giving you direct, explicit control over which local reference image gets used when generating or editing a picture. The tool also fixes a technical mismatch: the standard fallback command line version of Codex's image tool sends requests through the regular OpenAI API, which does not work correctly with certain image models tied to a Codex account login. This project instead routes requests through Codex's own base address, so those account linked models function as expected. Beyond reference image selection, the tool lets you choose which underlying request path is used, either the hosted image generation route or a separate image API route meant for generating and editing, using a command line flag. The built in version only supports one of these paths. You can also choose which specific image model to use and adjust how much reasoning effort it applies, rather than being stuck with whatever defaults Codex currently ships with. When you generate or edit an image, the original files and a log of requests and responses are saved in a folder inside your Codex configuration directory, and the finished images are then copied into your actual project folder, meaning the same image ends up stored in two places. To install it, you clone the project into your Codex skills folder and restart Codex. Once installed, you can trigger it from within Codex with a short natural language command, or you can run the included Python script directly from the command line for the same result.
An add-on for Codex's image generation skill that lets you explicitly choose your own reference image and image model.
Mainly Python. The stack also includes Python, Codex CLI, OpenAI API.
No license information is stated in the README, so usage rights are unclear.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
double-check against the repo, no cap.