Skip to content

3D-print softwareWorking MVP2026

Upload an STL. Get a price from an actual slice, not a guess from its bounding box.

42.7 g3h 12m€18.40
source of truth for price: the slicer
1
source of truth for price: the slicer
bounding-box guesses
0
bounding-box guesses
services: API, worker, web app, database
4
services: API, worker, web app, database

The problem

What people put up with

Most online print-quote tools multiply the size of a model's bounding box by a magic number. Hollow parts end up overpriced and dense parts underpriced, and the shop absorbs the difference. Doing it by hand means opening every file in a slicer.

The odd angle

The insight

Slicing is cheap and the slicer already knows the truth. We put a pinned OrcaSlicer behind a job queue and price from its output: filament grams, print time and the machine that can actually fit the part.

The approach

How we built it

  1. 1

    Pricing as a pure function

    The pricing engine is a separate package with no I/O: material cost, machine time, setup and quantity breaks in, a price out. The API and the slicer worker share it, so a quote and an invoice can never disagree.

  2. 2

    Build-volume checks first

    Before anything is sliced, the part is checked against each printer's build volume. Customers find out right away that a part is too big, not after a five-minute wait.

  3. 3

    A pinned slicer

    OrcaSlicer is pinned to one version in both the Docker image and the code. CLI flags change between releases, and a silent change in a flag means a silent change in price.

What it does

Features

  • STL upload with real slicing in a background worker
  • Material, infill, colour and quantity options
  • Stripe checkout and webhooks
  • Order tracking from paid to shipped
  • Self-hostable with Postgres, Redis and S3-compatible storage

The hard parts

What fought back

Slow work behind a fast interface

Slicing takes seconds to minutes. The interface shows progress and the queue retries with backoff, so a busy farm feels responsive instead of stuck.

What's next

On the roadmap

  • Several printers and materials per shop
  • Automatic plate packing for batch orders