AI sovereignty for the JVM
Quixotic AI provides a complete, open stack, for local AI on the JVM. From tokenizers and model formats, to a full inference engine with multi-modal capabilities.
No Python. No ONNX. No external services. Just AI, in a jar.
- Designed from first-principles for the JVM. AI runs end-to-end on the JVM. No sidecar servers, no ONNX, no IPC, no Python.
- Write once, accelerate everywhere. A common tensor API for CPUs and GPUs.
- Optional native acceleration. Fast matrix multiplication routines, competitive with
llama.cpp. - GraalVM's Native Image. First-class support for GraalVM Native Image: self-contained binaries, with small footprint and millisecond startup.
| Module | What it is | One-liner |
|---|---|---|
jinfer |
AI inference engine | Local AI inference for the JVM. Chat, vision, audio, embeddings, reranking, text-to-speech |
toknroll |
LLM tokenization | Token-perfect. Fast tokenizers for LLMs, pure Java, zero dependencies |
jam |
Quantized matrix multiplication | Just a matmul. Native implementations for several CPU ISAs |
jota |
Tensor engine | Write once, accelerate everywhere. Java, C, CUDA, HIP, Metal, OpenCL, Mojo |
gguf |
GGUF reader/writer | llama.cpp's model format, pure Java, zero dependencies |
safetensors |
Safetensors reader/writer | HuggingFace's model format, pure Java, zero dependencies |
Requires a JDK 25 and Maven 3.9; cmake and a C compiler to build the native jam kernels.
make test-fixtures # once after cloning: the tokenizer vocabularies and the enwik8 corpus
make test # the default suite: no models, no network
make ci # what a pull request runs: formatting, the suite, the corpus tests, the release shapemake help lists the rest.
The suites that need models or hardware are opt-in; each module's README describes how to run them.
See CONTRIBUTING.md before opening a pull request.