Optuna team releases Rustuna, a Rust-based, Python-dependency-free reimplementation of Optuna's hyperparameter optimizer

Turbovec is a Rust implementation of Google's TurboQuant, a quantization technique for vector search. It aims to improve performance and efficiency in similarity search.

Products·launched

Skeg: disk-first Rust vector DB for low memory and production reliability

Skeg is an open-source vector database written in Rust that prioritizes low memory usage and production reliability. It uses a disk-first approach where full vectors live on storage, with only small caches in memory, enabling sustained ingest, multi-tenancy, and high recall without the typical trade-offs. It solves the problem of memory pressure and performance degradation for developers running vector search in production.

Skeg is an open-source vector database written in Rust that prioritizes low memory usage and production reliability. It uses a disk-first approach where full vectors live on storage, with only small caches in memory, enabling sustained ingest, multi-tenancy, and high recall without the typical trade-offs. It solves the problem of memory pressure and performance degradation for developers running vector search in production.

Releases·Tool Release

runNburn: Rust GGUF runtime targets models larger than RAM

A developer announced runNburn, an open-source Rust inference runtime for quantized GGUF models that runs on CPU, CUDA, Metal, and Android, with experimental Vulkan and OpenCL paths. The runtime is designed to handle models that exceed available fast memory, addressing a gap in existing solutions.

A developer announced runNburn, an open-source Rust inference runtime for quantized GGUF models that runs on CPU, CUDA, Metal, and Android, with experimental Vulkan and OpenCL paths. The runtime is designed to handle models that exceed available fast memory, addressing a gap in existing solutions.

A developer posted Openheim on Hacker News, a runtime for running LLM agents across multiple providers, written in Rust. It matters to practitioners building agent applications who want a single runtime layer over different model vendors. Engagement was minimal, so community validation remains limited.

Sofka is a Rust-based terminal UI for Kubernetes, designed as a faster alternative to k9s. It integrates tightly with Flux, Helm, and External Secrets, and was largely coded by an LLM (Fable). It solves the problem of slow Kubernetes management for developers and operators.