Local AI Apps and Tools for Mac

MiniCPM is a family of compact language models for developers building local assistants and coding agents. It targets on-device workloads and hardware with limited computing resources.

JARLIS is a self-hosted AI email assistant for people who want help sorting their inbox while keeping control over outgoing replies. It prepares drafts for review, flags messages that need attention and archives low-priority mail. It never sends replies automatically.

Quicksand gives AI agents a Linux virtual machine for executing code and working with desktop apps on your own hardware. It's for developers who need a separate operating system for agent tasks, controlled through an async Python API.

Tel-Agent is an AI receptionist for businesses that want phone calls and customer messages handled by one agent on their own server. Recordings and transcripts stay on your machine, but its voice setup uses Deepgram, a cloud LLM and ElevenLabs, requiring internet access and your own API keys.

Whats Reader lets you revisit exported WhatsApp conversations and their attachments without sending the files to a service. It works with individual and group chats, so you can review old conversations or find details buried in messages and voice notes.

Talk DAT! is a free, open-source dictation app for people who want to turn speech into formatted writing across their desktop apps. It runs on Windows 10 and 11 and Apple silicon Macs with macOS 14 or later. Downloaded models let it work offline without an account, keeping speech and text on your computer.

Immersive Translate is a translation extension for people reading foreign-language articles, research and work documents. It keeps the original text beside the translation so readers can compare passages.

Multi-Agent Shogun runs and coordinates separate AI coding assistants on your computer or a Linux server. It's for developers dividing implementation and research across parallel tasks. Coordination happens locally, but Claude Code’s standard cloud connection sends prompts and model outputs over the network to their model provider; other agents use their chosen CLI's model access and authentication.

Blinko is an AI note app for people who want to capture passing thoughts and find them again later. Its self-hosted edition stores notes and data in your own environment.

OpenCrust is a self-hosted AI assistant platform for people who want agents to handle tasks through their messaging apps. Named agents can pass work to one another and retain conversation memory between sessions.

Helix is a paid, self-hosted platform for engineering teams coordinating coding agents. Agents work on separate tasks in parallel, while the team shares a view of their progress and reviews their work.

llm-fw inspects model requests for developers and teams using AI tools, SDKs and agents. It blocks suspected prompt injection and removes outgoing secrets before forwarding approved traffic to the model endpoint.

Deepchecks is a Python toolkit for data scientists and ML teams checking datasets and models before deployment and watching for problems in production. Its testing, CI and monitoring components reuse the same checks across those stages.

PyTorch Lightning is a Python framework for researchers and developers pretraining or fine-tuning AI models. It handles recurring training infrastructure while leaving the model logic in your PyTorch code.

RecurseChat is a Mac app for people who want private AI conversations and answers drawn from their own documents. Local models run without an internet connection or a separate server.

OpenBitFun is a desktop AI agent workspace for coding and other tasks that use files and shell commands on your machine or a selected SSH host. It runs on Windows, macOS and Linux. Model requests go to the configured provider; built-in cloud connections send them to external services, and custom providers are also supported.

PentAGI automates authorized penetration testing for security engineers, researchers and red teams. Its specialized AI agents turn a target and a plain-language goal into testing work you can follow through a web interface.

FastVLM is a family of local vision-language models for developers and researchers building applications that answer questions about images. It focuses on reducing the delay before an answer starts, especially with high-resolution images.

RisuAI is an AI roleplay client for people who want to create characters and fictional worlds, then develop stories through conversation. It supports chats with individual characters or groups.

Infinity is a search database for developers building retrieval-augmented generation (RAG), question answering and recommendation systems. It lets applications combine matches based on meaning with exact text matches and structured filters.

OpenCode is an AI coding agent for developers who want help writing code or exploring unfamiliar projects. You can choose a local model running on your hardware or connect a cloud provider.

LEANN is a local vector database for people building AI search over personal data or knowledge collections. It reduces index storage by computing embeddings on demand rather than keeping every embedding on disk.

AVCLabs Photo Enhancer AI is a proprietary desktop app for Windows and macOS that processes photo enhancements locally. It's aimed at photographers, people restoring family pictures, and creators preparing product images who want to apply corrections across a batch without uploading photos to a cloud service. AVCLabs also offers separate online image tools.

AgentSight is a local AI agent monitoring tool for developers investigating failed runs, slow steps or unexpected changes. It connects prompts and model calls to what agents actually do on the machine, including activity their own logs may miss.

ast-grep MCP lets AI coding assistants find code by syntax structure. It's for developers who need searches for specific constructs or relationships that text matching alone can't express.

Phone Harness connects an AI agent to a real iPhone or Android to carry out tasks through app screens. It's for developers who want phone automation without relying on an app API.

OpenUI is an AI UI builder for developers prototyping web components. Describe an interface in plain language, inspect the rendered result, and ask for changes before taking the code into a project.

OASIS is a social media simulator for researchers studying how information spreads, groups polarize, and users follow the crowd. It uses LLM agents as simulated participants in Twitter and Reddit environments.

ThinkRAG answers questions about your own documents using a knowledge base on your laptop. It's for researchers, students and other knowledge workers who want to find information across their files, with particular attention to Chinese text.

AI Infra Guard is a self-hosted AI security testing platform for developers and security teams checking agent workflows, model responses and AI infrastructure. It combines vulnerability scans and adversarial prompt tests in a web interface.

Selene is a desktop AI agent app for coding, browser tasks and personal assistants. You can choose local or cloud processing separately for chat, embeddings, voice and image generation.

VideoHighlighter is a desktop app for finding useful moments in long recordings without uploading footage for analysis. It helps creators cut highlights, reviewers inspect camera recordings, and interview editors locate material worth keeping.

Volyx Lens is an open-source desktop AI assistant for macOS, Windows and Linux that lets you choose which screen, microphone and system-audio inputs reach your AI provider. Linux support is experimental, with meeting audio captured on a best-effort basis. It's for people who want help with visible code, documents or meetings without continually copying context into a chat. A movable floating rail expands into chat and collapses when you don't need it.

Thuki is a floating AI assistant for Mac users who want help with the text or screen they're already viewing. It appears above other apps, including fullscreen windows, and automatically brings selected text into the conversation.