Minimalist ML framework for Rust
⑂ 1.7k◎ 909
Crate
cudarc
37
Repositories
97%
Actively maintained
34
Updated ≤ 30 days
62k
Total stars
34
Owners
Counted from the Cargo.toml manifests of the 37 indexed repositories that declare cudarc as a dependency — not download counts. Dependency data last verified 2026-08-13.
Often used with
Crates that show up unusually often in cudarc projects. The most distinctive pairings rank first — crates these projects use far more than the average indexed Rust project does, not just crates that are popular everywhere. Each percentage is the share of cudarc projects that also use it.
safetensors24.3%objc2-metal16.2%oneshot16.2%metal16.2%wide16.2%candle-nn21.6%candle-core24.3%half51.4%candle-transformers16.2%numpy18.9%hf-hub27%rustfft16.2%tokenizers35.1%ndarray43.2%rand_distr27%ort18.9%cc18.9%minijinja21.6%memmap251.4%parquet21.6%libloading35.1%arrow18.9%pyo343.2%rand_chacha21.6%
Used by these organizations
37 repositories · 34 pushed in the last 30 days
A Datacenter Scale Distributed Inference Serving Framework
⑂ 1.4k◎ 1.1k
ML-powered manga translator, written in Rust.
⑂ 332◎ 100
A blazing fast inference solution for text embeddings models
An extensible, state-of-the-art framework for columnar compression, and the fastest FOSS columnar file format. Formerly at @spiraldb, now an Incubation Stage project at LFAI&Data, part of the Linux Foundation.
A portable accelerated SQL query, search, and LLM-inference engine, written in Rust, for data-grounded AI apps and agents.
⑂ 217◎ 625
Tiny, no-nonsense, self-contained, Tensorflow and ONNX inference
⑂ 275◎ 146
Inference at the speed of light.
⑂ 218◎ 40
Multi-platform high-performance compute language extension for Rust.
⑂ 221◎ 156
Apache Mahout - an environment for quickly creating scalable, performant machine learning applications.
⑂ 996◎ 91
rvLLM: High-performance LLM inference in Rust. Drop-in vLLM replacement.
🦀 Low-level 3D Computer Vision library in Rust
⑂ 193◎ 84
Pure Rust + CUDA LLM inference engine
⑂ 98◎ 143
Pure Rust + CUDA LLM inference engine — no PyTorch, OpenAI-compatible, serves Qwen3 to Kimi-K2
⑂ 98◎ 143
Pure Rust Inference Engine
⑂ 97◎ 113
NextPlaid, ColGREP: Multi-vector search, from database to coding agents.
From-scratch Rust+CUDA inference engine, bit-exact by construction — NVFP4, MoE, MTP speculative decoding, tuned against measured limits of one RTX 5090 Laptop (sm_120a).
memra — from-scratch LLM inference engine for NVIDIA RTX 50-series (Rust + CUDA)
SciRS2 - Scientific Computing and AI in Rust., providing SciPy-compatible APIs while leveraging Rust's performance, safety, and concurrency features. Unlike traditional scientific libraries
⑂ 34◎ 2
High-performance KV cache storage for LLM inference — GPU offloading, SSD caching, and cross-node sharing via RDMA. Works with vLLM and SGLang.
KV cache control plane and evaluation platform for LLM inference
Rust-native GPU operator library for LLM inference, built with cuda-oxide
Camelid: a Rust-native local inference backend with evidence-gated model compatibility.
⑂ 26◎ 5
High-performance quantitative finance in Rust — 120+ stochastic processes, option pricing, calibration, fixed income, risk & copulas, with SIMD/GPU acceleration and Python bindings.
⑂ 8◎ 1
Rust inference package experiments
Accelerated Zero-knowledge Virtual Machine by Non-uniform Prover Based on GKR Protocol
⑂ 43◎ 97
Protein and molecule viewer, editor, simulator
ODE solver library in Rust
A machine learning library built with user convenience at its core
⑂ 5◎ 5
Next Generation Machine Learning, Statistics and Deep Learning in PURE Rust
⑂ 22◎ 23
A general-purpose extensible tensor computation library in Rust with CPU/GPU support
⑂ 3◎ 31
A Rust Library for High-Performance Tensor Exchange with Python
iris-mpc repository
Rust Implementation for WebNN
⑂ 8◎ 42
Minimalist, performant and auditable verifiable RISC-V vm written in Rust