Layer
Every printer in my comparison posts — the Bambu, the Prusa, the Creality — traces its lineage back to a university lab in Bath, England, and a mechanical engineering lecturer named Adrian Bowyer who, in 2005, started asking a strange question: could a 3D printer print most of its own parts? The project that question kicked off, RepRap, never became a household name the way the printers it inspired eventually did. But almost nothing about the desktop 3D printing industry as it exists today makes sense without understanding what RepRap actually set out to do and why it chose to give everything away for free.
The self-replicating idea
Bowyer's original pitch was closer to evolutionary biology than consumer electronics: build a machine that could reproduce by printing the plastic components needed to build a copy of itself. The metal rods, the motors, the electronics board — those still had to be bought — but the brackets, the gears, the structural frame pieces could all come out of the printer itself. A RepRap could, in theory, "breed," with one machine printing the parts for its successor. This is also where the name comes from: Replicating Rapid prototyper.
The self-replication framing was always more philosophy than daily practice — most RepRap owners never actually built a second machine from the first one's printed parts — but it shaped everything about the project's design decisions. If the goal is a machine anyone can reproduce, every part of the design has to be documented, buildable from commonly available materials, and free to copy.
Going open-source, on purpose
This is the decision that mattered most in the long run. Bowyer released the RepRap design under the GNU General Public License, meaning anyone could build one, modify it, and sell the modified version, as long as they kept the derivative open too. At the time, nearly all 3D printing patents and technology were locked up by a handful of industrial manufacturers selling machines that cost tens of thousands of dollars. RepRap's open license meant a hobbyist with a soldering iron and about $500 in parts could build something that, while far rougher, did fundamentally the same job.
Darwin and Mendel: the early machines
The first working RepRap, nicknamed "Darwin," was completed in 2007 and looked nothing like a modern printer — a boxy cubic frame of threaded rod and laser-cut plywood, with visible wiring and a build volume smaller than a shoebox. It was slow, finicky, and required real mechanical knowledge to keep running, but it printed, and the full designs were posted online for free.
Its successor "Mendel," released in 2009, is the machine most people picture when they think of "classic RepRap" — a distinctive wedge-shaped frame that became the visual ancestor of an entire generation of hobbyist printers. Mendel was easier to build, cheaper, and specifically designed with self-replication in mind: a higher percentage of its own parts could be printed on a Darwin or another Mendel than on the original Darwin design.
The companies that grew out of it
What happened next is the part that actually changed the market. Because the designs were open, nothing stopped someone from building a RepRap-derived machine, polishing the rough edges, and selling it fully assembled instead of as a kit of parts and instructions. MakerBot, founded in 2009 by three people connected to the RepRap community, did exactly that with its early Cupcake and Thing-O-Matic machines, and later Prusa Research grew out of Josef Prusa's own contributions to the RepRap project — his "Prusa Mendel" variant was, for years, the single most-built 3D printer design in the world, assembled by thousands of hobbyists from freely shared plans before Prusa Research ever became a company selling pre-built machines.
This pattern — open community design, forked and commercialized by a company, iterated on, sold — repeated itself across most of the industry's early history. Even machines that look nothing like a RepRap today, like a modern enclosed CoreXY printer, still owe structural ideas (the moving bed, the belt-driven gantry, the printed-part philosophy) to that original open lineage.
What RepRap got right
The core bet — that giving away the design would grow the overall market faster than protecting it would have — turned out to be correct almost beyond what anyone involved expected in 2005. A technology that had been the exclusive domain of industrial manufacturers selling $50,000 machines became something a hobbyist could build in a garage for a few hundred dollars within about five years, and something anyone could buy assembled for under $500 within about fifteen.
What's left of the original project today
The RepRap project itself, as an active hub, has faded from the center of attention it held in the late 2000s and early 2010s — most of the energy and most of the users moved to commercial machines and newer open firmware projects like Marlin and Klipper, which themselves descend from RepRap's original firmware work. The RepRap wiki and forums still exist, maintained by a smaller community than at its peak, mostly populated now by people building fully-printed machines in the project's original self-replicating spirit rather than people shopping for their first printer.
But the lineage is everywhere whether people know the name or not. Every time someone buys an enclosed CoreXY printer for under $500, they're buying the end point of a line that starts with a mechanical engineering lecturer in Bath asking whether a machine could print its own children.