Releases for the Magnitude coding agent

A benchmark for evaluating AI agents on realistic business workflows

Fast, AI-friendly CLI for Better Stack

Pure TypeScript git implementation: virtual filesystem client and embeddable server.

Elegant bindings for working with Git in your Node applications

A fast, helpful, and open-source document parser

Binary installation for rust projects

⚡ Rust/WebAssembly image processing library

Tile-Based Runtime for Ultra-Low-Latency LLM Inference

Intuitive, easy CLIs based on python type hints.

A GPU-rendered terminal emulator with inline 3D graphics 🐀🧀

Coding Agent singularly focused efficiency and context curation. Reduces API costs by 50-80% vs other agent AND improves the code quality at the same time. Uses Hash Anchored edits, massively paral…

agent multiplexer that lives in your terminal.

Beautiful git diff viewer, generate commits with AI, get summary of changes, all from the CLI

26m function call model that runs on incredibly small devices

💥 Fast State-of-the-Art Tokenizers optimized for Research and Production

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

cuda-oxide is an experimental Rust-to-CUDA compiler that lets you write (SIMT) GPU kernels in safe(ish), idiomatic Rust. It compiles standard Rust code directly to PTX — no DSLs, no foreign languag…

Gradient Bang is an online multiplayer universe where you explore, trade, battle, and collaborate with other players and with LLMs

Tree-based speculative decoding for Apple Silicon (MLX). ~10-15% faster than DFlash on code, ~1.5x over autoregressive. First MLX port with custom Metal kernels for hybrid model support.

Fast LLM speculative inference server for consumer hardware.

DFlash: Block Diffusion for Flash Speculative Decoding

DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm

On-device Speech AI for Apple Silicon

Open Source framework for voice and multimodal conversational AI