Skip to main content
OpenEng AI general availability · Q4 2026

The launch is close.
The platform is already real.

OpenEng AI reaches general availability in Q4 2026, when Pro, Enterprise, and pay-as-you-go open. Much of the platform isn’t a promise: the five engineering apps are live — Terminal, Kubernetes, and Data with published engines; Git and Chat with engines coming soon — along with ten public catalogs and the Console’s interface. Join the waitlist for launch-day access and updates.

No spam — one note when launch-day access opens. Submitting opens your email client to send us a quick note; or email us directly, or start now below.

Try the free apps Open the Console

Not a roadmap — live in the public catalogs today

1,092Agents
885Skills
754Templates
28On-device models
550Cells
1,524Roles
252Connectors
14Plugins
8Library bundles
6Global loops

Honest status

Live today, and what's next.

We keep a hard line between the two. Everything on the left is live now — the Git and Chat web apps load, but their engines are coming soon; everything on the right is either built and opening at launch, or on the roadmap. No vaporware in the 'live' column.

Live today

Live and verifiable right now

Five engineering apps, free, in beta

Terminal, Kubernetes, and Data are usable today: they run in your browser and drive a small engine on your own machine over a loopback-only, sealed channel. The Git and Chat web apps are live too; their engines are coming soon. Every app and every feature is free.
See the apps

Engines for Terminal, Kubernetes, and Data

One-line installers for Linux x86_64 (Kubernetes also on ARM64) and macOS; on Windows, run them under WSL. They update themselves: the Linux builds check releases.openeng.app, and the older macOS builds use the release service's former address until they're rebuilt. macOS builds are available for Terminal, Kubernetes, and Data; they are older builds pending a rebuild.
  • Newest builds: Terminal 0.0.4 · Kubernetes 0.0.5 · Data 0.0.4

10 catalogs, all live

Agents, skills, templates, models, cells, roles, connectors, plugins, library bundles, and loops — browse them today; adding them to a saved config opens with the Console's API.
  • 1,092 agents · 885 skills · 754 templates
  • 28 on-device models · 550 cells · 1,524 roles
  • 252 connectors · 14 plugins · 8 library bundles · 6 global loops

The Console

The Console interface is live at console.openeng.app: sign in and explore the config builder and the building blocks. Saving and publishing configs (an immutable id@version) opens with the Console's API; the Pro pages (API keys, schedulers, sandboxes, approval gates) unlock at launch.

One account, everywhere you sign in

One OpenEng identity across the five apps, the Console, and the Web Console, with Google sign-in in the browser. At launch, the AI engine starts with an API key from the Console (Pro).

What’s next

Opening at launch, and planned after it

Opens at launch · Q4 2026

The AI engine on Linux x86_64

The openeng binary is built and published: a Rust engine that runs on your machine, opens no network port, dials the OpenEng Broker, and owns the agent loop, models, memory, and every cloud call. It starts with a Console API key on a Pro plan, so it runs for you once Pro, the OpenEng AI API and the Broker open at launch.

The Web Console, driving your engine

app.openeng.app is the thin browser client for your AI engine. It reaches the engine through the Broker, and the engine's streamed replies are sealed end-to-end, so the Broker relaying them can't read them. Your open-weight models run on your machine.

Complexity routing — no modes

Every request is scored for complexity before the engine chooses to plan or to answer directly — by its built-in router, or by a model you bind. You never pick 'chat vs goal vs coding.'

Pro, Enterprise, and pay-as-you-go

Paid plans open for purchase at launch, and prices are announced then. Trial terms are announced at launch. See the plans.

On the roadmap

Git and Chat engines

Both web apps are live at git.openeng.app and chat.openeng.app. Their engines are coming soon; until they are published, neither has an installer on this site.

More platforms

macOS rebuilds of the newest app engines are pending; until then, macOS stays one version behind. The AI engine is published for Linux x86_64 only, and no engine has a native Windows build yet — on Windows, use WSL.

Full per-worker OS isolation

Today, where bubblewrap works on the host, a kernel sandbox confines the shell commands the agent runs (without it they run unconfined), and the model daemon runs in an enclosure: a systemd resource cap where a user systemd session works, and a read-only root where bubblewrap works. Extending hard OS-level isolation to every individual worker is the next hardening milestone.

Embedded media generation

Today's on-device models cover chat, coding, planning, review, and grounding. On-device image and media generation, enclosed like every built-in model, is on the roadmap.

We ship each piece when it’s proven, not before. Q4 2026 is the only date on this page, and it is for OpenEng AI’s paid plans; the engineering apps are already free.

You don't have to wait

The live half is enough to start.

The engineering apps are free and live now, and the Console's interface is open to explore — no waitlist required. The waitlist is for the GA note and launch-day access to OpenEng AI.

step 1 · apps

Install an app engine

One line installs the Terminal, Kubernetes, or Data engine on Linux x86_64 (Kubernetes also ARM64) or macOS (WSL on Windows). Open the app, sign in, and run the one-time command it gives you.

step 2 · console

Explore the Console

Sign in at console.openeng.app and explore the config builder and the building blocks: models, agents, skills, and templates. Saving and publishing a config (an immutable id@version) opens with the Console's API.

step 3 · launch

Join the waitlist

Get the GA note and launch-day access — or just see the pricing first.

Start free — the Terminal engine

Every engine and platform, including the AI engine, is on the download page.

Be there on day one.

Join the waitlist for launch-day access and the GA announcement — or start with the free engineering apps right now.

Explore the free apps Read the AI docs