Professional / Tools

Exocortex MediaForge

Tool share

Exocortex MediaForge

A local Windows desktop command deck for turning prompt lists into organized OpenAI or local ComfyUI batches.

If you already have prompt lists and just want to run batches locally instead of babysitting prompts in a browser, this is for that.

The packaged download is for Windows. If you want to run it on macOS, the repo now includes a simple source-run setup path too.

Packaged for Windows, with a source-run path for macOS.
OpenAI mode uses your own OPENAI_API_KEY.
Local ComfyUI image or video workflows are supported too.
Curated local presets include SDXL, FLUX, Chroma, Wan, and HunyuanVideo.
Writes images or videos, manifests, and logs to disk.
No accounts, no telemetry, no hosted dashboard.

More routes

What it does

Practical batch workflow

Paste prompts or load `.txt`, `.csv`, or `.json`, choose an output folder, switch between OpenAI or local ComfyUI, and run a batch locally.

The app saves images or MP4 videos directly to disk, keeps filenames clean, writes a `manifest.json` and `generation_log.csv`, and lets you retry failed items without rerunning the whole list.

It is built for practical asset generation, not for adding another hosted layer around a workflow that already lives in folders. The local side now includes a few curated ComfyUI presets for popular image and video model families too.

Why it is useful

Batch runs instead of one-off prompts.

Retry failed items without restarting the whole batch.

Keep category folders, manifests, and logs together.

Use the same prompt files you already work with for stills or clips.

Setup

1. Download the Windows ZIP.

2. Run the included first-time setup helper.

3. Add your own `OPENAI_API_KEY` for OpenAI mode, or point it at ComfyUI for local mode.

4. If you use ComfyUI, choose a preset or manual workflow path and point it at a compatible API workflow.

5. Pick an output folder and run a batch.

6. On macOS, run the included shell helper scripts from the repo instead of the Windows ZIP.

Billing truth

This app does not include credits. OpenAI image and video usage are billed through your own OpenAI API account, while ComfyUI local mode uses your own local machine and workflow setup.

The point here is just a cleaner desktop workflow around the generation APIs, not a new billing layer.

Supported formats

TXT

One prompt per line. Good when you just want to paste a batch and run it.

CSV

Use columns like prompt, asset_id, category, filename_prefix, background, size, or quality when you want more structure.

JSON

Useful if your prompt planning already lives in a small pipeline or generator script.

FAQ

Does it use my own API key?

OpenAI mode uses your own OPENAI_API_KEY. Local ComfyUI mode runs against your own local setup instead, so it does not need an OpenAI key.

Does it keep output local?

The surrounding workflow stays local: prompt files, output folders, manifests, and logs all live on your machine. Generation requests still go to OpenAI because that is how the APIs work.

What gets saved?

Generated images or MP4 videos go to the output folder you choose, plus a manifest.json and generation_log.csv so you can inspect what happened later.

Does it support video too?

Yes. You can switch the app between image mode and video mode. OpenAI video works directly, and local ComfyUI can batch saved video workflows too if your API-format workflow writes a real clip file.