--- title: "Benchmarks - BuildPrint wiki" description: "Frame rate, loading time and freezes of BuildPrint, Litematica and HoloBuilder on the same builds, from small to 11 million blocks." url: https://voxelsprout.com/buildprint/wiki/benchmarks --- [BuildPrint](https://voxelsprout.com/buildprint) [Wiki](https://voxelsprout.com/buildprint/wiki)Benchmarks # Benchmarks Frame rate, loading time and freezes of BuildPrint, Litematica and HoloBuilder on the same builds, from small to 11 million blocks. These are the results of a benchmark I run in the game itself. It generates the same test builds every time, loads them into each mod in turn and measures from the same spot. Only one mod has a build loaded at a time. ## Frame rate Average fps, with the 1% low in brackets. Higher is better. **Vanilla** is the same spot with nothing loaded. HoloBuilder draws only part of each build (see below), so its numbers are marked partial and are not like for like. | Test | Blocks | Vanilla | BuildPrint | Litematica | HoloBuilder (partial) | | ----------------- | ------ | --------- | ---------- | ---------- | --------------------- | | Every block state | 32k | 292 (150) | 226 (117) | 103 (62) | 49 (17) | | Generated town | 106k | 276 (149) | 215 (107) | 224 (88) | 11 (8) | | Sparse world | 200k | 297 (150) | 230 (80) | 110 (54) | 24 (21) | | Solid stone | 1M | 275 (145) | 235 (121) | 242 (110) | 9 (7) | | Complex geometry | 1M | 276 (148) | 53 (12) | 32 (28) | 5 (4) | | Redstone parts | 1M | 272 (141) | 75 (43) | 28 (16) | 4 (4) | | Generated town | 1.1M | 277 (147) | 225 (82) | 52 (34) | 12 (10) | | Sparse world | 2M | 309 (149) | 89 (36) | 73 (47) | 14 (12) | | Complex geometry | 5M | 277 (139) | 15 (9) | 7 (7) | 5 (4) | | Redstone parts | 5M | 272 (148) | 19 (11) | 7 (7) | 4 (3) | | Solid stone | 10M | 269 (148) | 225 (110) | 96 (57) | 9 (8) | | Generated town | 11.2M | 391 (157) | 186 (76) | 45 (13) | 10 (9) | ## Loading Seconds until the first part of the build shows, seconds until all of it is drawn, and the longest single frame while loading. A long frame is a moment the game stops. | Test | BuildPrint | Litematica | HoloBuilder (partial) | | ---------------------- | ---------------------- | ----------------------- | --------------------- | | Every block state, 32k | 0.2 s / 0.4 s, 40 ms | 0.4 s / 0.4 s, 134 ms | 0.3 s / 0.3 s, 58 ms | | Generated town, 106k | 0.1 s / 0.2 s, 34 ms | 0.3 s / 0.3 s, 142 ms | 0.1 s / 0.1 s, 123 ms | | Sparse world, 200k | 0.3 s / 1.0 s, 18 ms | 1.4 s / 1.4 s, 50 ms | 0.3 s / 0.3 s, 56 ms | | Solid stone, 1M | 0.1 s / 0.5 s, 11 ms | 0.5 s / 0.5 s, 190 ms | 0.1 s / 0.1 s, 125 ms | | Complex geometry, 1M | 0.2 s / 1.1 s, 65 ms | 5.3 s / 5.3 s, 4.1 s | 0.1 s / 0.1 s, 350 ms | | Redstone parts, 1M | 0.2 s / 0.8 s, 65 ms | 3.0 s / 3.0 s, 2.1 s | 0.2 s / 0.2 s, 319 ms | | Generated town, 1.1M | 0.2 s / 0.5 s, 47 ms | 1.8 s / 1.8 s, 618 ms | 0.3 s / 0.3 s, 104 ms | | Sparse world, 2M | 2.2 s / 6.2 s, 93 ms | 6.1 s / 6.1 s, 87 ms | 2.4 s / 2.4 s, 114 ms | | Complex geometry, 5M | 0.7 s / 12.9 s, 237 ms | 27.3 s / 27.3 s, 19.6 s | 0.5 s / 0.5 s, 643 ms | | Redstone parts, 5M | 0.8 s / 7.1 s, 170 ms | 29.1 s / 29.1 s, 19.9 s | 0.5 s / 0.5 s, 1.4 s | | Solid stone, 10M | 0.3 s / 4.5 s, 16 ms | 10.4 s / 10.4 s, 10.3 s | 0.4 s / 0.4 s, 850 ms | | Generated town, 11.2M | 1.7 s / 2.0 s, 33 ms | 10.0 s / 10.0 s, 1.2 s | 3.3 s / 3.3 s, 130 ms | ## What the numbers say * On small, plain builds BuildPrint and Litematica are about the same. In the 106k town and the 1M stone box Litematica is a few percent ahead, which is within the difference between two runs. * On bigger everyday builds BuildPrint keeps a much higher frame rate: 225 against 52 fps in the 1.1M town, 186 against 45 in the 11.2M town. * Very dense builds, 5 million blocks of stairs, panes, plants or redstone parts where almost no face is hidden, are slow in every mod. That is the graphics card drawing that much, not the mod. * These numbers are BuildPrint drawing the whole build, like Litematica. With default settings BuildPrint shortens how far the ghost draws when the frame rate drops, so in play the dense builds run at a higher frame rate than shown here. * BuildPrint reads files in the background and adds the build bit by bit, so the game keeps running while it loads. Litematica reads on the game thread, and the game stopped for up to 20 seconds on the biggest tests. * On the dense 5M tests BuildPrint shows the first part within a second, but takes 7 to 13 seconds to draw all of it. * HoloBuilder draws at most 65,536 blocks per frame, and only within 96 blocks of you by default. None of these builds fits in that, so it drew only part of each one: on the 5M tests about 1% of the blocks. It also draws every block again every frame, which is why its frame rate is low even then. ## The tests | Test | What it is | | ----------------- | ---------------------------------------------------------------------------------------------------------------------- | | Every block state | Every block in every state the game has, one per spot. Checks that every model draws. | | Generated town | Plots with houses, yards, trees and redstone rooms, the same mix of blocks at every size. The closest to a real build. | | Sparse world | Blocks scattered through a large box, 512 × 96 × 512 or 1024 × 192 × 1024. | | Solid stone | One solid box of stone. | | Complex geometry | Stairs, fences, walls, glass, panes, plants and lanterns in random states. | | Redstone parts | Observers, repeaters, comparators, pistons, wire and torches in random states. | ## How it was tested * MacBook Pro, Apple M4 Max, 36 GB, macOS. Minecraft 26.2 with Fabric, render distance 24, no other performance mods. October 2026. * BuildPrint 0.1.19, Litematica 0.28.8 with MaLiLib, HoloBuilder 0.1.5\. All on their default settings, except that BuildPrint’s lag guard was off, so it drew the whole build like Litematica. * All three mods were installed in one game. Per test: nothing loaded first, then each mod twice, in the order A, B, C, C, B, A, 10 seconds each, so heat and memory over the run affect every mod the same. * The camera stands just outside the build, a bit above it, looking in, so the build fills the screen. In the sparse test it stands in the middle. * Graphics card use was not measured: macOS does not report it per app. One machine, one run, generated builds. Your numbers will be different, and a different graphics card can change the dense tests the most. [Previous Coming from Litematica](https://voxelsprout.com/buildprint/wiki/from-litematica) [Next Bug reports](https://voxelsprout.com/buildprint/wiki/bug-reports) Something missing or wrong on this page? Say so on [Discord](https://discord.gg/CvdkGbEW4y).