Resources
Guides, comparisons & use cases
Start with a hub, drill into one topic, or browse the full index on the site map — without wading through walls of links.
- Comparisons
14 honest matchups vs Cursor, Copilot, Claude Code, and more.
- Setup guides
Install on Windows, macOS, Linux — OpenRouter, Ollama, and env setup.
- Use cases
Privacy, refactors, Obsidian memory, browser automation, and workflows.
Popular comparisons
All 14 comparisons →- vs CursorCompare Liminal AI and Cursor for coding: free fair-source agent vs AI IDE. Local tools, any model, full tool trace — honest 2026 comparison.
- vs GitHub CopilotLiminal AI vs GitHub Copilot: autonomous local agent with tools vs inline suggestions. Compare pricing, privacy, and use cases.
- vs Claude CodeCompare Liminal AI and Claude Code: fair-source local harness vs Anthropic's terminal agent. Models, cost, and control.
- vs ChatGPTLiminal AI vs ChatGPT for software work: local agent that edits your repo vs cloud chat. Tools, privacy, and when to use each.
- vs ClineLiminal AI vs Cline (Claude Dev): standalone fair-source agent vs VS Code extension. Tools, models, memory, and pricing.
- vs WindsurfCompare Liminal AI and Windsurf (Codeium): local fair-source agent vs AI-powered IDE. Models, privacy, pricing.
Setup guides
All 7 guides →- Install Liminal on WindowsInstall the Liminal AI coding agent on Windows 10/11 with PowerShell. Node.js 22+, API key, and first run in minutes.
- Install Liminal on macOSInstall Liminal AI on macOS with one curl command. Apple Silicon and Intel supported. Node 22+ required.
- Install Liminal on LinuxInstall Liminal AI on Ubuntu, Debian, Fedora, and other Linux distros. One script, local agent, any LLM.
- OpenRouter + LiminalPoint Liminal AI at OpenRouter for DeepSeek, Claude, GPT, and more. API key, model slugs, and harness env vars in five minutes.
Use cases
All 8 use cases →- Local-first AI codingRun an AI coding agent on your machine: session traces, tool execution, and memory on disk you control. Optional cloud LLMs or fully local inference.
- Refactor large codebasesPlan and execute large refactors with repo maps, grep, semantic tools, multi-file edits, tests, lint self-heal, and optional dynamic workflows across modules.
- Obsidian + agent memoryConnect Liminal to Obsidian: vault_write, hybrid BM25+vector recall, memory graph, workspace-scoped notes, and curator-safe pruning — a coding agent with a second brain you own.
- Browser automation for devsHeadless Playwright in the agent loop: navigate, snapshot, act, extract — for JS-heavy docs, QA repros, and web-backed dev tasks with stealth and session limits.
From the blog
All posts →- How to run an AI coding agent without your own API keyMost AI coding agents make you create a provider account and paste an API key before you can do anything. Liminal AI gives you both options: bring your own key, or use managed inference and start in one sign-in.
- Cline and Roo Code alternatives — when to use Liminal insteadVS Code agent extensions are great in-editor. Here is when a standalone fair-source harness with workflows and web UI wins.
- Tabnine vs an autonomous coding agent — completions are not the whole jobTabnine optimizes inline suggestions. Liminal runs full repo tasks with git, tests, and browser tools. How to choose in 2026.
Technical reference: docs.vireondynamics.com/liminal/Full site map