Libraries

A protobuf code-generation framework for Rust, built at Dropbox for its storage and sync systems, offering Rust-specific extensions and zero-copy deserialization that general-purpose protobuf libraries do not provide.

• Libraries

A small Swift library that calculates sunrise and sunset times, including civil, nautical, and astronomical twilight, entirely on-device from a date and a location, with no network access required.

• Libraries

A Python library that generates high-performance CPU and GPU code for array-based computations, such as dense linear algebra and PDE solvers, by transforming a simple description under the user's control.

• Libraries

A Python package that compiles high-level symbolic descriptions of stencil computations, such as finite-difference methods, into optimized, parallel code for CPUs, GPUs, and clusters.

• Libraries

A lightweight, pure-Swift library for reading and validating Apple In-App Purchase receipts locally, without sending them to a server.

• Libraries

A Julia framework that automatically parallelizes code across CPU cores, GPUs, and multiple machines by scheduling tasks as a directed-acyclic graph, without manual coordination.

• Libraries
• Data engineering

A model-based-design autopilot system for building control software for drones, cars, ships, and robots, pairing a C/C++ embedded firmware with a MATLAB/Simulink simulation and algorithm library.

• Libraries

An open-source Python platform that uses large language models to automatically design and optimize algorithms, with unified interfaces across methods, tasks and models plus a graphical interface for running experiments.

• Libraries

A set of Swift macros that generates a type's Codable implementation, covering flattened nested keys, fallback keys, default values, custom coding strategies, and untagged enums, without hand-written boilerplate.

• Libraries

A modular Swift SDK for running text-to-speech, speech-to-text, voice activity detection, and speech-to-speech models on Apple Silicon through MLX, with importable modules for just the models an app needs.

• Speech
• Libraries

A Swift package for building tools and applications with large language models and vision-language models using MLX Swift, with LoRA and full fine-tuning support.

• Models
• Libraries

A Swift library for running large language models locally on macOS, iOS, watchOS, tvOS and visionOS, built around llama.cpp with automatic chat-template handling from a model's GGUF metadata.

• Inference
• Libraries

An interactive GUI for arranging and annotating matplotlib figures for publication, which records every change as generated Python code that reproduces the finished figure.

• Libraries

A Java library for communicating with Minecraft clients and servers over its network protocol, letting developers build custom bots, clients or servers.

• Libraries

A palette-quantization library that converts RGBA images into palette-based 8-bit indexed images, including the alpha channel, for producing small PNGs and GIFs.

• Libraries

A Swift package that is an API-compatible drop-in replacement for Apple's Foundation Models framework, adding support for Core ML, MLX, llama.cpp and several hosted model APIs behind the same interface.

• Inference
• Libraries

An open-source remote procedure call system designed for multichip embedded systems and heterogeneous multicore SoCs, using plain C for remote functions and a small code footprint.

• Libraries

A fast JSON Schema validator for JavaScript that compiles a schema into generated validation code instead of interpreting it at runtime.

• Libraries

FlashRank is an ultra-lightweight Python library that reranks search results before they are fed to an LLM, using cross-encoder and listwise LLM-based rerankers small enough to run on a CPU with no PyTorch or Transformers dependency.

• RAG
• Libraries

FastEmbed-rs is a Rust library for generating text and image vector embeddings and for reranking search results locally, using ONNX Runtime for inference and Hugging Face tokenizers for encoding, with synchronous use and no dependency on Tokio.

• RAG
• Libraries

Optimum Quanto is a PyTorch quantization backend for Hugging Face's optimum library, now in maintenance mode, that inserts quantization and dequantization automatically and supports int2, int4, int8 and float8 weights and activations across CPU, CUDA and MPS devices.

• Models
• Libraries

Ollama-rs is a Rust client library for the Ollama API, covering text completion including streaming, chat, embeddings, function calling and model management, built by following Ollama's own API documentation.

• Inference
• Libraries

A lightweight, YAML-configured Python robot simulator for navigation, control and learning research, with built-in collision detection for robots, sensors and environments.

• Libraries

The official Go client library for the Claude API, providing access to Anthropic's language models from Go applications.

• Libraries