Transformer inference via unconventional compute: a single algebraic primitive, a tag-system scheduler, and Apple's Neural Engine through convolution.

Fast tokenizer for language models, compatible with SentencePiece, Tokenizers, Tiktoken and more. Supports BPE, Unigram and WordPiece tokenization in JavaScript, Python and Rust.

Library for reducing tail latency in RAM reads

Developer-friendly OSS embedded retrieval library for multimodal AI. Search More; Manage Less.

Python bindings for access to the on-device model at the core of Apple Intelligence through the Foundation Models framework

Open-Source Frontier Voice AI

A 10x faster and more accurate text wrapping util in < 2KB (min)

Vulkan bindings for Rust

rvLLM: High-performance LLM inference in Rust. Drop-in vLLM replacement.

Native macOS Wayland Compositor written in Rust using Smithay. Experience seamless Linux app streaming on macOS without XQuartz.

Fast, accurate & comprehensive text measurement & layout

A faster than tiktoken tokenizer with first-class support for Vercel's AI SDK.

A pure-Python implementation of the Nvidia CuTe layout algebra intended to be approachable and easy to learn.

An experiment in durability

Grafeo is a pure-Rust, high-performance graph database that can be embedded as a library or run as a standalone database, with optional in-memory or persistent storage. Grafeo supports both LPG and…

Miryoku is an ergonomic, minimal, orthogonal, and universal keyboard layout.

Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"

Official repo for BWLer: Barycentric Weight Layer

Hyperbolic Graph Convolutional Networks in PyTorch.

Renovation Engine for DirectX Games

GPU-accelerated terminal board that puts all your sessions on an infinite canvas

Reference code for the Slug Algorithm.

Depix is a PoC for a technique to recover plaintext from pixelized screenshots.

High-performance Vulkan driver with automated resource management and execution

Oxyde ORM is a type-safe, Pydantic-centric asynchronous ORM with a high-performance Rust core designed for clarity, speed, and reliability.

Decentralized DB. Remember Everything. Everywhere. All At Once.