OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Infrastructure & Serving

Llama3.Java

Llama3.Java is a single-file Java implementation for inference with Llama 3, 3.1 and 3.2 models. It parses GGUF model files and provides command-line chat and instruct modes.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Llama3.Java — OpenViST project profile in AI Infrastructure & Serving.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you want a compact, Java-based route to local Llama inference and can supply a supported GGUF model file.

Who it is for

  • Java developers who want to run supported Llama models from the command line.
  • Learners studying practical LLM inference in Java.
  • Developers exploring a Java-based inference component for an application using Llama models.

Documented uses

  • Run supported Llama models using GGUF weights.
  • Interact with models in chat or instruct mode.
  • Build a JAR or GraalVM Native Image executable.
  • Experiment with Java Vector API and JVM compiler optimizations for inference.
Before you choose

Requirements & limits

  • Java 21 or later is required, including for the MemorySegment memory-mapping feature.
  • The implementation supports Llama 3, 3.1 and 3.2, with listed weight formats and quantizations.
  • GraalVM Native Image is optional; the Makefile includes JAR and native build targets.
  • The README recommends GraalVM 25 or later for best performance.
  • The project is a single Java file with no dependencies and is licensed under MIT.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Build a native executable with a preloaded GGUF model

Create a GraalVM Native Image executable that preloads a specific GGUF model.

Before you start

  • Java 21 or later
  • GraalVM and the repository's Makefile
  • A GGUF model file

Steps

  1. Set PRELOAD_GGUF to the path of the model file.

  2. Run make native.

Expected outcome

A larger native executable is generated with that model preloaded; it can also run other models, with the usual parsing overhead.

Guide limitations

  • The README does not provide model-specific performance guarantees.
  • GraalVM 25 or later is recommended for best performance.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
14 October 2026
Original website title
mukel/llama3.java · Official homepage title · retained in its source language
How we review sourcesSuggest a correction