Open Source Mac LLM Inference Tools
Sort By:
LLM Inference Tools for Mac
View 17 business solutionsBrowse free open source LLM Inference tools and projects for Mac below. Use the toggles on the left to filter open source LLM Inference tools by OS, license, language, programming language, and project status.
-
Full-stack observability with actually useful AI | Grafana CloudBuilt on open standards like Prometheus and OpenTelemetry, Grafana Cloud includes Kubernetes Monitoring, Application Observability, Incident Response, plus the AI-powered Grafana Assistant. Get started with our generous free tier today.
-
Custom VMs From 1 to 96 vCPUs With 99.95% UptimeLive migration and automatic failover keep workloads online through maintenance. One free e2-micro VM every month.
-
1
whisper.cpp
Port of OpenAI's Whisper model in C/C++
whisper.cpp is a lightweight, C/C++ reimplementation of OpenAI’s Whisper automatic speech recognition (ASR) model—designed for efficient, standalone transcription without external dependencies. The entire high-level implementation of the model is contained in whisper.h and whisper.cpp. The rest of the code is part of the ggml machine learning library. The command downloads the base.en model converted to custom ggml format and runs the inference on all .wav samples in the folder samples. whisper.cpp supports integer quantization of the Whisper ggml models. Quantized models require less memory and disk space and depending on the hardware can be processed more efficiently. -
2
GPT4All
Run Local LLMs on Any Device. Open-source
GPT4All is an open-source project that allows users to run large language models (LLMs) locally on their desktops or laptops, eliminating the need for API calls or GPUs. The software provides a simple, user-friendly application that can be downloaded and run on various platforms, including Windows, macOS, and Ubuntu, without requiring specialized hardware. It integrates with the llama.cpp implementation and supports multiple LLMs, allowing users to interact with AI models privately. This project also supports Python integrations for easy automation and customization. GPT4All is ideal for individuals and businesses seeking private, offline access to powerful LLMs. -
3
llama.cpp
Port of Facebook's LLaMA model in C/C++
The llama.cpp project enables the inference of Meta's LLaMA model (and other models) in pure C/C++ without requiring a Python runtime. It is designed for efficient and fast model execution, offering easy integration for applications needing LLM-based capabilities. The repository focuses on providing a highly optimized and portable implementation for running large language models directly within C/C++ environments. -
4
Open WebUI
User-friendly AI Interface
Open WebUI is an extensible, feature-rich, and user-friendly self-hosted AI platform designed to operate entirely offline. It supports various LLM runners like Ollama and OpenAI-compatible APIs, with a built-in inference engine for Retrieval Augmented Generation (RAG), making it a powerful AI deployment solution. Key features include effortless setup via Docker or Kubernetes, seamless integration with OpenAI-compatible APIs, granular permissions and user groups for enhanced security, responsive design across devices, and full Markdown and LaTeX support for enriched interactions. Additionally, Open WebUI offers a Progressive Web App (PWA) for mobile devices, providing offline access and a native app-like experience. The platform also includes a Model Builder, allowing users to create custom models from base Ollama models directly within the interface. With over 156,000 users, Open WebUI is a versatile solution for deploying and managing AI models in a secure, offline environment. -
Gemini 3 and 200+ AI Models on One PlatformBuild generative AI apps with Vertex AI. Switch between models without switching platforms.
-
5
ONNX Runtime
ONNX Runtime: cross-platform, high performance ML inferencing
ONNX Runtime is a cross-platform inference and training machine-learning accelerator. ONNX Runtime inference can enable faster customer experiences and lower costs, supporting models from deep learning frameworks such as PyTorch and TensorFlow/Keras as well as classical machine learning libraries such as scikit-learn, LightGBM, XGBoost, etc. ONNX Runtime is compatible with different hardware, drivers, and operating systems, and provides optimal performance by leveraging hardware accelerators where applicable alongside graph optimizations and transforms. ONNX Runtime training can accelerate the model training time on multi-node NVIDIA GPUs for transformer models with a one-line addition for existing PyTorch training scripts. Support for a variety of frameworks, operating systems and hardware platforms. Built-in optimizations that deliver up to 17X faster inferencing and up to 1.4X faster training. -
6
LocalAI
The free, Open Source alternative to OpenAI, Claude and others
LocalAI is an open-source platform that allows users to run large language models and other AI systems locally on their own hardware. It acts as a drop-in replacement for APIs such as OpenAI, enabling developers to build AI-powered applications without relying on external cloud services. The platform supports a wide range of model types, including text generation, image creation, speech processing, and embeddings. LocalAI can run on consumer-grade hardware and does not necessarily require a GPU, making it accessible for local development and private deployments. It integrates with multiple backends like llama.cpp, transformers, and diffusers to support different AI workloads. With its self-hosted architecture and OpenAI-compatible API, LocalAI enables developers to build secure, local-first AI applications. -
7
vLLM
A high-throughput and memory-efficient inference and serving engine
vLLM is a fast and easy-to-use library for LLM inference and serving. High-throughput serving with various decoding algorithms, including parallel sampling, beam search, and more. -
8
ncnn
High-performance neural network inference framework for mobile
ncnn is a high-performance neural network inference computing framework designed specifically for mobile platforms. It brings artificial intelligence right at your fingertips with no third-party dependencies, and speeds faster than all other known open source frameworks for mobile phone cpu. ncnn allows developers to easily deploy deep learning algorithm models to the mobile platform and create intelligent APPs. It is cross-platform and supports most commonly used CNN networks, including Classical CNN (VGG AlexNet GoogleNet Inception), Face Detection (MTCNN RetinaFace), Segmentation (FCN PSPNet UNet YOLACT), and more. ncnn is currently being used in a number of Tencent applications, namely: QQ, Qzone, WeChat, and Pitu. -
9
Gitleaks
Protect and discover secrets using Gitleaks
Gitleaks is a fast, lightweight, portable, and open-source secret scanner for git repositories, files, and directories. With over 6.8 million docker downloads, 11.2k GitHub stars, 1.7 million GitHub Downloads, thousands of weekly clones, and over 400k homebrew installs, gitleaks is the most trusted secret scanner among security professionals, enterprises, and developers. Gitleaks-Action is our official GitHub Action. You can use it to automatically run a gitleaks scan on all your team's pull requests and commits, or run on-demand scans. If you are scanning repos that belong to a GitHub organization account, then you'll have to obtain a license. Gitleaks can be installed using Homebrew, Docker, or Go. Gitleaks is also available in binary form for many popular platforms and OS types on the releases page. In addition, Gitleaks can be implemented as a pre-commit hook directly in your repo or as a GitHub action using Gitleaks-Action. -
Fully Managed MySQL, PostgreSQL, and SQL ServerCloud SQL handles your database ops end to end, so you can focus on your app.
-
10
OpenVINO
OpenVINO™ Toolkit repository
OpenVINO™ is an open-source toolkit for optimizing and deploying AI inference. Boost deep learning performance in computer vision, automatic speech recognition, natural language processing and other common tasks. Use models trained with popular frameworks like TensorFlow, PyTorch and more. Reduce resource demands and efficiently deploy on a range of Intel® platforms from edge to cloud. This open-source version includes several components: namely Model Optimizer, OpenVINO™ Runtime, Post-Training Optimization Tool, as well as CPU, GPU, MYRIAD, multi device and heterogeneous plugins to accelerate deep learning inferencing on Intel® CPUs and Intel® Processor Graphics. It supports pre-trained models from the Open Model Zoo, along with 100+ open source and public models in popular formats such as TensorFlow, ONNX, PaddlePaddle, MXNet, Caffe, Kaldi. -
11
OnnxStream
Lightweight inference library for ONNX files, written in C++
The challenge is to run Stable Diffusion 1.5, which includes a large transformer model with almost 1 billion parameters, on a Raspberry Pi Zero 2, which is a microcomputer with 512MB of RAM, without adding more swap space and without offloading intermediate results on disk. The recommended minimum RAM/VRAM for Stable Diffusion 1.5 is typically 8GB. Generally, major machine learning frameworks and libraries are focused on minimizing inference latency and/or maximizing throughput, all of which at the cost of RAM usage. So I decided to write a super small and hackable inference library specifically focused on minimizing memory consumption: OnnxStream. OnnxStream is based on the idea of decoupling the inference engine from the component responsible for providing the model weights, which is a class derived from WeightsProvider. A WeightsProvider specialization can implement any type of loading, caching, and prefetching of the model parameters. -
12
DeepSpeed
Deep learning optimization library: makes distributed training easy
DeepSpeed is an easy-to-use deep learning optimization software suite that enables unprecedented scale and speed for Deep Learning Training and Inference. With DeepSpeed you can: 1. Train/Inference dense or sparse models with billions or trillions of parameters 2. Achieve excellent system throughput and efficiently scale to thousands of GPUs 3. Train/Inference on resource constrained GPU systems 4. Achieve unprecedented low latency and high throughput for inference 5. Achieve extreme compression for an unparalleled inference latency and model size reduction with low costs DeepSpeed offers a confluence of system innovations, that has made large scale DL training effective, and efficient, greatly improved ease of use, and redefined the DL training landscape in terms of scale that is possible. These innovations such as ZeRO, 3D-Parallelism, DeepSpeed-MoE, ZeRO-Infinity, etc. fall under the training pillar. -
13
LLaVA
Visual Instruction Tuning: Large Language-and-Vision Assistant
Visual instruction tuning towards large language and vision models with GPT-4 level capabilities. -
14
Transformer Engine
A library for accelerating Transformer models on NVIDIA GPUs
Transformer Engine (TE) is a library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper GPUs, to provide better performance with lower memory utilization in both training and inference. TE provides a collection of highly optimized building blocks for popular Transformer architectures and an automatic mixed precision-like API that can be used seamlessly with your framework-specific code. TE also includes a framework-agnostic C++ API that can be integrated with other deep-learning libraries to enable FP8 support for Transformers. As the number of parameters in Transformer models continues to grow, training and inference for architectures such as BERT, GPT, and T5 become very memory and compute-intensive. Most deep learning frameworks train with FP32 by default. This is not essential, however, to achieve full accuracy for many deep learning models. -
15
Hello AI World
Guide to deploying deep-learning inference networks
Hello AI World is a great way to start using Jetson and experiencing the power of AI. In just a couple of hours, you can have a set of deep learning inference demos up and running for realtime image classification and object detection on your Jetson Developer Kit with JetPack SDK and NVIDIA TensorRT. The tutorial focuses on networks related to computer vision, and includes the use of live cameras. You’ll also get to code your own easy-to-follow recognition program in Python or C++, and train your own DNN models onboard Jetson with PyTorch. Ready to dive into deep learning? It only takes two days. We’ll provide you with all the tools you need, including easy to follow guides, software samples such as TensorRT code, and even pre-trained network models including ImageNet and DetectNet examples. Follow these directions to integrate deep learning into your platform of choice and quickly develop a proof-of-concept design. -
16
DeepCamera
Open-Source AI Camera. Empower any camera/CCTV
DeepCamera empowers your traditional surveillance cameras and CCTV/NVR with machine learning technologies. It provides open-source facial recognition-based intrusion detection, fall detection, and parking lot monitoring with the inference engine on your local device. SharpAI-hub is the cloud hosting for AI applications that helps you deploy AI applications with your CCTV camera on your edge device in minutes. SharpAI yolov7_reid is an open-source Python application that leverages AI technologies to detect intruders with traditional surveillance cameras. The source code is here It leverages Yolov7 as a person detector, FastReID for person feature extraction, Milvus the local vector database for self-supervised learning to identify unseen persons, Labelstudio to host images locally and for further usage such as label data and train your own classifier. It also integrates with Home-Assistant to empower smart homes with AI technology. -
17
Genv
GPU environment management and cluster orchestration
Genv is an open-source environment and cluster management system for GPUs. Genv lets you easily control, configure, monitor and enforce the GPU resources that you are using in a GPU machine or cluster. It is intended to ease up the process of GPU allocation for data scientists without code changes. -
18
huggingface_hub
The official Python client for the Huggingface Hub
The huggingface_hub library allows you to interact with the Hugging Face Hub, a platform democratizing open-source Machine Learning for creators and collaborators. Discover pre-trained models and datasets for your projects or play with the thousands of machine-learning apps hosted on the Hub. You can also create and share your own models, datasets, and demos with the community. The huggingface_hub library provides a simple way to do all these things with Python. -
19
Diffusers
State-of-the-art diffusion models for image and audio generation
Diffusers is the go-to library for state-of-the-art pretrained diffusion models for generating images, audio, and even 3D structures of molecules. Whether you're looking for a simple inference solution or training your own diffusion models, Diffusers is a modular toolbox that supports both. Our library is designed with a focus on usability over performance, simple over easy, and customizability over abstractions. State-of-the-art diffusion pipelines that can be run in inference with just a few lines of code. Interchangeable noise schedulers for different diffusion speeds and output quality. Pretrained models that can be used as building blocks, and combined with schedulers, for creating your own end-to-end diffusion systems. We recommend installing Diffusers in a virtual environment from PyPi or Conda. For more details about installing PyTorch and Flax, please refer to their official documentation. -
20
LLamaSharp
C#/.NET binding of llama.cpp, including LLaMa/GPT model inference
The C#/.NET binding of llama.cpp. It provides APIs to infer the LLaMa Models and deploy it on the local environment. It works on both Windows, Linux and MAC without the requirement for compiling llama.cpp yourself. Its performance is close to llama.cpp. Furthermore, it provides integrations with other projects such as BotSharp to provide higher-level applications and UI. -
21
ChatGLM.cpp
C++ implementation of ChatGLM-6B & ChatGLM2-6B & ChatGLM3 & GLM4(V)
ChatGLM.cpp is a C++ implementation of the ChatGLM-6B model, enabling efficient local inference without requiring a Python environment. It is optimized for running on consumer hardware. -
22
ChatLLM.cpp
Pure C++ implementation of several models for real-time chatting
chatllm.cpp is a pure C++ implementation designed for real-time chatting with Large Language Models (LLMs) on personal computers, supporting both CPU and GPU executions. It enables users to run various LLMs ranging from less than 1 billion to over 300 billion parameters, facilitating responsive and efficient conversational AI experiences without relying on external servers. -
23
KServe
Standardized Serverless ML Inference Platform on Kubernetes
KServe provides a Kubernetes Custom Resource Definition for serving machine learning (ML) models on arbitrary frameworks. It aims to solve production model serving use cases by providing performant, high abstraction interfaces for common ML frameworks like Tensorflow, XGBoost, ScikitLearn, PyTorch, and ONNX. It encapsulates the complexity of autoscaling, networking, health checking, and server configuration to bring cutting edge serving features like GPU Autoscaling, Scale to Zero, and Canary Rollouts to your ML deployments. It enables a simple, pluggable, and complete story for Production ML Serving including prediction, pre-processing, post-processing and explainability. KServe is being used across various organizations. -
24
KubeAI
Private Open AI on Kubernetes
Get inferencing running on Kubernetes: LLMs, Embeddings, Speech-to-Text. KubeAI serves an OpenAI compatible HTTP API. Admins can configure ML models by using the Model Kubernetes Custom Resources. KubeAI can be thought of as a Model Operator (See Operator Pattern) that manages vLLM and Ollama servers. -
25
RWKV Runner
A RWKV management and startup tool, full automation, only 8MB
RWKV (pronounced as RwaKuv) is an RNN with GPT-level LLM performance, which can also be directly trained like a GPT transformer (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, fast training, saves VRAM, "infinite" ctxlen, and free text embedding. Moreover it's 100% attention-free. Default configs has enabled custom CUDA kernel acceleration, which is much faster and consumes much less VRAM. If you encounter possible compatibility issues, go to the Configs page and turn off Use Custom CUDA kernel to Accelerate.
