Free browser-based 2.5D CAD/CAM with nesting and trochoidal clearing

Oct 31, 2026 Last reply: 17 hours ago 4 Replies
Jump to replies (4)

Most hobby CAM runs out of heavyweight desktop suites. Against that backdrop, a free, open-source 2.5D CAD/CAM package that lives in the browser feels like a fresh take. PureCutCNC just shipped version 0.5.0, with a desktop build available too, aimed at hobby CNC routing: draw a design, and the app generates the toolpaths.



The nesting stands out. Parts are laid out on the stock in quantities you set, spaced according to the cutter that will cut them, and the layouts keep clear of clamps. Small parts can even sit inside the holes of bigger ones, and stock can be an irregular outline, with cutting boards given as the example. Pocket clearing runs trochoidal orbits ring by ring rather than plowing a full-width path, the rings link at depth instead of retracting between passes, and the same approach carries over to 3D roughing.



Other choices I rate highly: arc lead-ins and lead-outs bring the cutter onto walls along tangent arcs, avoiding witness marks beside the wall, and edge routes helix down at every level rather than plunging vertically. A CAM Plan reads the drawn geometry and proposes operations and tools, but nothing is created until you have reviewed and edited the proposal. In my view that gate is a sensible safeguard.


3D finishing comes as parallel, waterline and constant-scallop strategies, with the constant-scallop mode driven by a scallop height you set instead of a stepover, and any variation can be split by slope ranges; the maker cites 0-30 degrees for parallel and 30-90 for waterline. STEP import and text following a circle or an arbitrary path round things out, along with performance work and bug fixes.

One open item: there is no generic Fanuc post processor yet, and plasma support does not exist while the maker learns that process. The status bar still reads beta, and treating this as a work in progress seems fair.



Project by PureCutCNC. Watch it on their YouTube channel:

formatting link


The clamp avoidance being built into toolpath generation rather than flagged after the paths exist is the choice I would single out. Flagging afterwards usually leaves you manually patching g-code or re-running everything, while generating around the clamps from the start should give output you can run as-is. The trade-off, as I see it, is that the generator is only as safe as the clamp positions you feed it. Still, catching collisions at generation time looks like the right place for it.

How does somebody actually set up an irregular stock outline for the nesting? Is the shape sketched directly inside the app, or does it come from converting existing geometry somehow? I have been wondering about that since the cutting-board example is one of the showcased uses for squeezing parts onto odd stock.

That one has an answer: you draw the outline as a shape, then right-click it and pick the option to use it as stock. From there the nesting can squeeze parts into whatever that outline happens to be, which is exactly the route behind the cutting-board demonstration.

If the maker wants a direction for an upcoming cycle, fourth-axis indexing looks like the obvious candidate. The notes are already sitting in the log, and the maker has said indexing itself would be relatively easy; the holdup is only that the initial scope is undefined. Picking a narrow first slice, such as basic indexing for round parts, would give users something real to test soon instead of waiting on a full 4-axis implementation.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required