AgentOS chat window showing a streaming assistant response with sidebar and input bar
Founders price · first 200 buyers

The coding agent that lives on your Mac.

AgentOS runs MLX, llama.cpp, LM Studio, and EXO clusters — your code, your models, your machine. No subscriptions. No cloud round-trips. Just an agent loop wired to the models you already trust.

Get AgentOS $99 See how it works ↓
  • Apple Silicon only
  • Offline-first
  • 2 active Macs per license
  • Lifetime updates
Projects

Ground every chat in your code.

Point AgentOS at a folder. It picks up your code, MEMORY.md, and DECISIONS.md so every conversation starts grounded — not from scratch.

  • Per-project context that survives across sessions
  • Project-bound chats with their own history
  • Memory file injection on every turn — no manual paste
AgentOS Projects landing page with a centered folder icon and Start a project call-to-action
Skills

183 built-in skills. No marketplace.

Battle-tested recipes ship with the app — from "scaffold a SwiftUI macOS app with a working .xcodeproj" to "diagnose a Core Data threading violation." Attach them per conversation, and the library is refreshed every release so it tracks frontier model capability — never goes stale.

  • Bundled — no skill marketplace to maintain
  • One-click attach from the input bar
  • Refreshed every release to match frontier model capability
  • Add your own as Markdown files in ~/Library

Auto-attach by intent (skills surfaced automatically when the conversation needs them) ships in a future release — paired with a toggle so users on M1 / 16 GB Macs can opt out, since expanding context every turn slows load times on older hardware.

AgentOS Skills browser showing the bundled skill library with the macOS SwiftUI recipe open
Models

Curated models that actually tool-call.

Most local models look great in benchmarks and fall over the first time you ask them to call a tool. We curate the ones that work — Qwen 3.6, Gemma 4, gpt-oss, MiniMax — and ship them as one-click downloads sized for your Mac's RAM.

  • Pre-vetted GGUFs sized to your unified memory
  • Active model picker right in the chat input bar
  • Split-file support for the big ones (480B, 120B)
AgentOS Models pane showing the curated library and available downloads
Backends

Four backends. One protocol.

llama.cpp ships bundled and auto-starts. LM Studio plugs in with one click. EXO lets you fan inference across multiple Macs on your network. The Local API Server exposes everything as OpenAI-compatible so Xcode 16 Intelligence and any other client just work.

  • llama.cpp · Metal-tuned, split-file support, bundled
  • LM Studio · drop-in OpenAI endpoint
  • EXO · cluster inference across Mac Studios + MBPs
  • Local API Server · works with Xcode AI assistant
AgentOS Connection settings showing the four-backend picker with llama.cpp active and ready

One time. Yours forever.

AgentOS is a $99 founders-price license while the first 200 spots last — then it goes to $199. No tiers, no subscriptions, no usage caps.

Founders · 200 spots
Founders price
$99 one time
Lifetime license. Locked in even after the price goes up.
  • All current and future features
  • 2 active Macs per license
  • Lifetime updates via Sparkle
  • Direct email support from Max
  • Priority on feature requests during v1.x
Claim founders price →
Standard
$199 one time
After the first 200 founders licenses sell.
  • All current and future features
  • 2 active Macs per license
  • Lifetime updates via Sparkle
  • Standard email support
Join the waitlist

Be first in line.

We'll email you when the founders-price gates open. No newsletter, no marketing blasts — just the launch ping.

We'll only email you once — when the founders price opens.