OL
Ollama
FeaturedLocal AI
Run open models locally with one command. Simplest local AI runtime for macOS, Windows and Linux.
✓ macOS✓ Windows✓ Linux
LLMLocal AICLIAPI
$ ollama run qwen2.5
Open Source • Free • Local-first
Discover. Install. Run Open-Source AI.
Explore open-source AI models, agents, frameworks and local AI tools with verified installation commands for macOS, Windows and Linux.
Find an AI → Understand it → Choose OS → Install → Run → Build
Search AI models, agents, frameworks…
Ollama, Qwen, DeepSeek, Llama, OpenClaw, CrewAI, LangChain, llama.cpp and more.
⌕
19 results
LL
Llama
AI Model
Meta's open-weight LLM family. The reference open model for local and self-hosted inference.
✓ macOS✓ Windows✓ Linux
LLMCodingOpen Model
$ ollama run llama3.2
QW
Qwen
FeaturedAI Model
Alibaba's open-weight LLM family. Strong multilingual and coding performance at many sizes.
✓ macOS✓ Windows✓ Linux
LLMCodingMultilingual
$ ollama run qwen2.5
DE
DeepSeek
AI Model
DeepSeek's open-weight reasoning models. Cheap API plus open local weights.
✓ macOS✓ Windows✓ Linux
LLMReasoningCoding
$ ollama run deepseek-r1:7b
GE
Gemma
AI Model
Google's open-weight model family, built from Gemini research. Efficient on consumer hardware.
✓ macOS✓ Windows✓ Linux
LLMEfficientGoogle
$ ollama run gemma2:9b
MI
Mistral
AI Model
Mistral AI's open-weight models. Fast and compact, popular for local and edge inference.
✓ macOS✓ Windows✓ Linux
LLMFastEdge
$ ollama run mistral
LL
llama.cpp
Local AI
C++ inference engine for open models. Extremely lightweight, runs on laptops and CPUs.
✓ macOS✓ Windows✓ Linux
InferenceC++CPUGGUF
$ llama-cli -m model.gguf
OP
OpenHands
FeaturedAI Agent
AI software development agent that writes, executes and debugs code in a sandboxed environment.
✓ macOS✓ Windows✓ Linux
Coding AgentAutonomousDocker
$ docker run -it -p 3000:3000 ghcr.io/all-hands-ai/openhands
OP
OpenCode
FeaturedAI Agent
Open-source AI coding agent built for the terminal and editor. Runs locally with your CLI and models.
✓ macOS✓ Windows✓ Linux
Coding AgentCLITerminalLocal
$ opencode
AI
Aider
AI Agent
AI pair programming in your terminal. Works with many local and cloud models, fully open source.
✓ macOS✓ Windows✓ Linux
Coding AgentCLIPair Programming
$ aider
OP
OpenJarvis
AI Agent
A local-first AI home assistant that can send messages, control apps and more using Java.
✓ macOS✓ Windows✓ Linux
Personal AIAssistantLocalJava
$ java -jar openjarvis.jar
OP
OpenClaw
AI Agent
An open-source personal AI agent with memory, skills and Claude integration for your desktop.
✓ macOS✓ Windows✓ Linux
Personal AIMemorySkillsDesktop
$ openclaw
AU
AutoGPT
AI Agent
The original autonomous agent platform. Build, deploy and run continuous AI agents.
✓ macOS✓ Windows✓ Linux
AutonomousAgentPlatform
$ npx autogpt
LA
LangChain
Framework
The most popular framework for building LLM applications, agents and tool integrations.
✓ macOS✓ Windows✓ Linux
FrameworkLLMPythonAgents
$ pip install langchain
CR
CrewAI
Framework
Framework for orchestrating role-playing autonomous AI agents to collaborate on tasks.
✓ macOS✓ Windows✓ Linux
FrameworkMulti-agentPython
$ pip install crewai
SM
smolagents
Framework
Hugging Face's minimal agent framework. Write agents in a few lines of Python.
✓ macOS✓ Windows✓ Linux
FrameworkAgentsPythonHF
$ pip install smolagents
LL
LlamaIndex
RAG
The data framework for LLM apps. Built for RAG workloads over your own documents.
✓ macOS✓ Windows✓ Linux
RAGFrameworkDataPython
$ pip install llama-index
VL
vLLM
Local AI
High-throughput, memory-efficient LLM serving engine. Best for GPU servers.
✓ Linux
ServingGPUHigh-performance
$ pip install vllm
LO
LocalAI
Local AI
Open-source, self-hosted, OpenAI-compatible local API. Runs models on CPU and GPU.
✓ macOS✓ Windows✓ Linux
Local AIOpenAI-compatibleSelf-hosted
$ docker run -p 8080:8080 quay.io/go-skynet/local-ai
Featured AI Tools
Hand-picked open-source projects, ready to install today.
🧠
AI Models
Llama, Qwen, DeepSeek & more
5 available →
🤖
AI Agents
Coding & personal agents
6 available →
🛠️
Agent Frameworks
LangChain, CrewAI & more
3 available →
💻
Local AI
Ollama, llama.cpp & more
4 available →
📚
RAG
LlamaIndex & retrieval
1 available →
🔌
MCP
Model Context Protocol
0 available →
OL
Ollama
FeaturedLocal AI
Run open models locally with one command. Simplest local AI runtime for macOS, Windows and Linux.
✓ macOS✓ Windows✓ Linux
LLMLocal AICLIAPI
$ ollama run qwen2.5
QW
Qwen
FeaturedAI Model
Alibaba's open-weight LLM family. Strong multilingual and coding performance at many sizes.
✓ macOS✓ Windows✓ Linux
LLMCodingMultilingual
$ ollama run qwen2.5
OP
OpenHands
FeaturedAI Agent
AI software development agent that writes, executes and debugs code in a sandboxed environment.
✓ macOS✓ Windows✓ Linux
Coding AgentAutonomousDocker
$ docker run -it -p 3000:3000 ghcr.io/all-hands-ai/openhands
OP
OpenCode
FeaturedAI Agent
Open-source AI coding agent built for the terminal and editor. Runs locally with your CLI and models.
✓ macOS✓ Windows✓ Linux
Coding AgentCLITerminalLocal
$ opencode
AI Directory
Models, coding agents, personal agents, frameworks and local runtimes — all in one place.
OP
OpenHands
FeaturedAI Agent
AI software development agent that writes, executes and debugs code in a sandboxed environment.
✓ macOS✓ Windows✓ Linux
Coding AgentAutonomousDocker
$ docker run -it -p 3000:3000 ghcr.io/all-hands-ai/openhands
OP
OpenCode
FeaturedAI Agent
Open-source AI coding agent built for the terminal and editor. Runs locally with your CLI and models.
✓ macOS✓ Windows✓ Linux
Coding AgentCLITerminalLocal
$ opencode
AI
Aider
AI Agent
AI pair programming in your terminal. Works with many local and cloud models, fully open source.
✓ macOS✓ Windows✓ Linux
Coding AgentCLIPair Programming
$ aider
OP
OpenJarvis
AI Agent
A local-first AI home assistant that can send messages, control apps and more using Java.
✓ macOS✓ Windows✓ Linux
Personal AIAssistantLocalJava
$ java -jar openjarvis.jar
OP
OpenClaw
AI Agent
An open-source personal AI agent with memory, skills and Claude integration for your desktop.
✓ macOS✓ Windows✓ Linux
Personal AIMemorySkillsDesktop
$ openclaw
AU
AutoGPT
AI Agent
The original autonomous agent platform. Build, deploy and run continuous AI agents.
✓ macOS✓ Windows✓ Linux
AutonomousAgentPlatform
$ npx autogpt
AI Agent Directory
Not one generic category — coding, personal and autonomous agents, each with its own install flow.
Coding Agents
Write, debug and refactor code for you
OpenCodeGitHub ↗
Open-source AI coding agent built for the terminal and editor. Runs locally with your CLI and models.
Coding AgentCLITerminal
Platforms: macOS · Windows · Linux
OpenHandsGitHub ↗
AI software development agent that writes, executes and debugs code in a sandboxed environment.
Coding AgentAutonomousDocker
Platforms: macOS · Windows · Linux
AiderGitHub ↗
AI pair programming in your terminal. Works with many local and cloud models, fully open source.
Coding AgentCLIPair Programming
Platforms: macOS · Windows · Linux
Personal Agents
Local assistants with memory and skills
OpenJarvisGitHub ↗
A local-first AI home assistant that can send messages, control apps and more using Java.
Personal AIAssistantLocal
Platforms: macOS · Windows · Linux
OpenClawGitHub ↗
An open-source personal AI agent with memory, skills and Claude integration for your desktop.
Personal AIMemorySkills
Platforms: macOS · Windows · Linux
Autonomous Agents
Self-directed task execution
Run an agent locally with Ollama models — ollama run qwen2.5 + Aider/OpenCode = a private coding agent on your laptop.
Ollama Hub
Ollama is the easiest way to run open models locally. It powers local AI agents, coding tools and RAG pipelines with one command.
Install Ollama
Downloads and runs the official Ollama installer script for macOS.
curl -fsSL https://ollama.com/install.sh | sh✓ Verify — ollama --version
ollama --versionPrints the installed version.
Windows: run irm https://ollama.com/install.ps1 | iex in PowerShell instead.
Llama
GeneralRuns locally with Ollama ⚡
ollama run llama3.2Pulls and runs the model in your terminal.
Qwen
General / CodingRuns locally with Ollama ⚡
ollama run qwen2.5Pulls and runs the model in your terminal.
Gemma
EfficientRuns locally with Ollama ⚡
ollama run gemma2:9bPulls and runs the model in your terminal.
DeepSeek
ReasoningRuns locally with Ollama ⚡
ollama run deepseek-r1:7bPulls and runs the model in your terminal.
Mistral
FastRuns locally with Ollama ⚡
ollama run mistralPulls and runs the model in your terminal.
gpt-oss
CodingRuns locally with Ollama ⚡
ollama run gpt-oss:20bPulls and runs the model in your terminal.
Ollama also exposes an OpenAI-compatible /v1 API on port 11434, so agents and frameworks like LangChain, CrewAI and Aider can use your local models.
Run AI Locally
Pick the runtime that fits your hardware. Requirements are verified per project — we don’t promise it will run on any machine.
| Runtime | Platform | Primary use | License | Notes |
|---|---|---|---|---|
| OllamaGitHub ↗ | macOS / Windows / Linux | Easy local AI | MIT | One-command install and model manager |
| llama.cppGitHub ↗ | macOS / Windows / Linux | Lightweight inference | MIT | C++ engine, runs on almost any CPU |
| vLLMGitHub ↗ | Linux | High-performance serving | Apache 2.0 | GPU servers, high throughput |
| LocalAIGitHub ↗ | macOS / Windows / Linux | OpenAI-compatible local API | MIT | Drop-in replacement for OpenAI endpoints |
Hardware requirements vary by model size and quantization. Always check the official documentation before installing.
Installation Demo
Pick a tool and your operating system. We explain every command — not just show it.
About
Ollama
Run open models locally with one command. Simplest local AI runtime for macOS, Windows and Linux.
LLMLocal AICLIAPI
GitHub ↗Ollama — macOS install
Last verified: September 2026- 1
Install Ollama
curl -fsSL https://ollama.com/install.sh | shDownloads and runs the official Ollama installer script for macOS.
- 2
Run a model
ollama run qwen2.5Pulls and starts the Qwen 2.5 model in your terminal.
✓ Verify installation
Run this to confirm everything installed correctly:
ollama --versionPrints the installed version.
What does the main command mean?
$ ollama run qwen2.5
ollamaOllama CLI — the local AI runtimerunPulls and starts the named modelqwen2.5Argument passed to the command above.The Discover → Install → Run Loop
The installation layer for open-source AI — not just another listing site.
1
Find an AI
Search models, agents, frameworks and local runtimes.
2
Understand it
Read what it does, its license, hardware and verified status.
3
Choose OS
Pick macOS, Windows or Linux — instructions adjust automatically.
4
Get verified commands
Copy install + verify commands sourced from official docs.
5
Install
Step-by-step guide that explains each command, not just shows it.
6
Run & Build
Launch the model or agent and start building.
Submit an open-source AI project
Built or using a great open-source model, agent or runtime? Submit it so the community can install and run it locally.
Every install command is reviewed and labeled with its verification date before publishing.