Skip to main content

How MOD Works

MOD turns heavy file-processing into something you can draw: drop your files in, wire up a pipeline of processing steps on a canvas, hit run, and MOD farms the work out to powerful machines and hands you the results. No servers to manage, no GPU to own.

This page is the 2-minute mental model. Once it clicks, the rest of the product is obvious.

The six pieces

ConceptIn one line
File (cog)Anything you upload — an image, mesh, video, audio clip, document. Lives in your Data area.
NodeOne processing step ("remove background", "generate a normal map", "transcribe audio"). MOD ships 130+.
WorkflowA pipeline of nodes wired together on the canvas. It's versioned, so you can iterate safely.
RunOne execution of a workflow version against your files. Goes pending → running → succeeded.
WorkerThe machine that actually does the work. Use MOD's shared pool for free, or enroll your own.
CreditWhat a run costs. Your plan includes a monthly allowance; each node has a small credit cost.

That's the whole model: files → nodes → workflow → run → (workers) → results, paid for in credits.

How a job flows

You Dashboard / Builder Workers
─── ─────────────────── ───────
upload a file ─► stored as a "cog"
build a workflow ► nodes wired on a canvas
click Run ─► run created ──────────────► a worker picks up each
node's task, runs it in a
progress + logs ◄────────── container, returns output
download result ◄─ output cogs

You watch progress live (per-node bars, streaming logs); MOD handles scheduling the tasks onto whatever hardware fits.

Two places you'll work

  • Dashboard — everything around workflows: your files, billing, team, settings, analytics, the marketplace. Lives at your main MOD URL.
  • Workflow Builder — the visual canvas where you build and run workflows. Opens in its own tab from the Dashboard sidebar.

The same things are available over a REST API, so anything you do by clicking, you can automate.

What you can build

MOD is general-purpose file processing, but it's especially strong at visual/3D work. A taste:

  • Background removal & image cleanup — cut-outs, deblurring, exposure balancing.
  • PBR texture generation — turn a photo into albedo / normal / roughness / AO / height maps.
  • 3D & photogrammetry — reconstruct 3D from photos (Colmap, Gaussian splatting, NeRF), convert and decimate meshes.
  • Video & audio — transcode, split, remove objects from video, text-to-speech.
  • AI/LLM nodes — classification, segmentation (SAM2), depth estimation, LLM-assisted analysis.
  • AEC — BIM/IFC conversion, compliance and variance checks for the built environment.
  • Your own code — wrap any script, Docker image, or repo into a reusable node.

See the full Node Library for the catalog.

Start here

  1. Getting Started — sign up, upload, build, run (5 min).
  2. Tutorial: Your First Workflow — a complete worked example (remove a background), in the UI and the API.
  3. Workflow Builder Guide — nodes, connections, batch mode.
  4. API Reference — automate everything, with a live "try it" explorer.

New to the vocabulary? The Glossary defines every term in plain English.