Community listing page, reviews may not be monitored by the author.
Velox
No reviews yet
Velox takes the lag out of dropped items.
Community voices
Reviews
No reviews yet. Be the first to review this project!
Get it on
Available Platforms
About
Project Details
For authors
Embed Badge
If you're the author of this project, you can embed a live badge anywhere that supports HTML or Markdown. It updates automatically whenever ratings change.
Use HTML for any page that supports it, or Markdown for README files and Markdown-based descriptions.
Identifiers
Platform IDs
About
Description

Velox is built to take work off your server when lots of dropped items are lying around. Big explosions, mob drops piling up, a storage room emptied onto the floor: instead of working out where each item goes one by one, Velox does them all in one go.
⚡ What it does
- 🧺 Dropped items in one batch. Items are worked out together instead of one by one, so a crowded floor costs the server far less time per tick.
- 👀 Looks the same to players. Items on the ground stay where they are and move the way they always did.
- 🌊 Vanilla where it matters. Items in water, lava, on hoppers or on slabs keep the normal game's behaviour.
- 🖥️ Server only. Players do not need to install anything.
Velox runs its physics on Celeris, which must be installed on the server too.
🏁 How fast
Milliseconds per server tick. The budget for 20 ticks per second is 50 ms.
- 🖥️ 20,000 falling items (Minecraft 1.21.1, dedicated NeoForge server, CI runner, median of 3 runs on 30 Oct 2026): 28.0 ms instead of 493.0 ms, 18× faster. Resting on the ground: 21.3 ms instead of 260.4 ms, 12× faster. (The normal game moves a lot between runs, so read the factor as good to about 20 %.)
- 🧹 8,000 resting items: 8.7 ms instead of 54.9 ms, 6× faster. 2,000 resting items: 1.3 ms instead of 8.5 ms, 7× faster.
- ⚙️ Java, JNI and FFM kernels tie. Velox runs on any of Celeris's three physics engines and they give the same tick to within noise (20,000 falling items: 28.5 ms Java, 28.0 ms JNI, 31.2 ms FFM; the whole physics step is 1–2 % of the server thread with a native kernel and about 7 % with the Java one). The gain over the normal game comes from Velox skipping Minecraft's per-item work, not from native code.
- 🧱 Crowded piles gain most. In the local tests, 100,000 items packed on 224 × 224 blocks: 171 ms instead of 3,315 ms, 19× faster.
- 🏗️ A million items: 2.8 s per tick instead of 25.5 s. Velox does not make a million items playable, but it makes a big drop survivable.

The benchmarks page has every number, the test machine and how each run was made. Results vary by about 20 % between runs on a shared machine.
📦 Requirements
- 🟩 Minecraft: 1.21 to 1.21.11 and 26.1 to 26.3.
- 🔶 Loader: NeoForge, on the server.
- 🧩 Mod: Celeris 0.3.1.1 or newer, on the server.
- ☕ Java: 21 for 1.21.x, 25 for 26.x.
📚 Learn more
The Velox wiki has the settings, how it works, the benchmarks and the build notes.
License
| Content | License |
|---|---|
| Source code | AGPL-3.0, see LICENSE-AGPL |
| Artwork, branding and documentation | CC BY-NC-SA 4.0, see LICENSE-CC |
Source code: you may study, modify and redistribute it under the AGPL; if you run a modified version as a network service, its source must be available to that service's users.
Artwork: credit Bichal, no commercial use without permission, and share derivatives under the same license.
See LICENSE for the full summary.
Code: AGPL-3.0 · Art: CC BY-NC-SA 4.0 · Made by Bichal
Screenshots
Gallery
Versions
Files
Relations
Project Relations
More like this
Similar Mods
Suggestions use data such as tags, dependencies, dependents, descriptions, titles, and more to rank how much they overlap with this mod.
On ModDex
Community snapshot
By the numbers
Statistics
Get it on
Available Platforms
On ModDex
Community snapshot
By the numbers