Skip to content

One platform · engineering apps + AI

Build with OpenEng

OpenEng is one platform with two pillars: five free engineering apps that run in your browser and execute on a small engine on your own machine (the Git and Chat engines are coming soon), and OpenEng AI, an agentic ecosystem for running open-weight models on your machine by default. One account works across all of it.

Engineering apps

OpenEng is a suite of five professional developer apps — Terminal, Kubernetes, Data, Git, and Chat — that run in your web browser but execute on a small native engine you run on your own machine. Terminal, Kubernetes and Data run on one today; the Git and Chat engines are coming soon. They are free: every app, every feature.

Five apps that drive your real infrastructure from the browser, with the work done on your machine. A thin web UI talks to a sealed local engine on loopback, so your code, credentials and data go only where you point the engine (your own shells, clusters, databases and repositories), never through OpenEng. Chat is the app that runs open-weight models: by default from your disk, so prompts stay on your machine unless you point a model at an external server.

Terminal

A reliable local shell with command blocks, smart suggestions, SSH and cloud shells.

127.0.0.1:50052

Kubernetes

Clusters, namespaces, pods, logs, exec, events and nodes — on local clusters, or on cloud clusters through a connector.

127.0.0.1:50054

Data

Query SQL, NoSQL, graph, key-value and object-storage databases, directly or through a connector in your own cloud.

127.0.0.1:50057

Git

Repositories, status, diffs, commits, branches, remotes, a commit graph and CI & CD runs. The engine is not published yet.

127.0.0.1:50053

Chat

Run local models, create groups with system prompts, and let one agent act across Terminal, Kubernetes and Git. The engine is not published yet.

127.0.0.1:50056

App engines: Terminal and Data on Linux x86_64 and macOS; Kubernetes on Linux x86_64 and ARM64 and macOS; Windows through WSL. The macOS builds are older releases pending a rebuild. The Git and Chat web apps are live, and their engines are not published yet. See Install, platforms and updates.

Connectors

Bring-your-own-cloud, source-only functions that let an app reach cloud resources over HTTPS — your credentials and data stay in your account.

One local architecture

Every app is a thin UI over a sealed loopback engine. Local-first and private, and the wire has no field for an OpenEng secret or product prompt.

Get started with an app

Install the Terminal engine, pair it with the web app using a one-time key, and run your first command.

OpenEng AI

OpenEng AI is an agentic ecosystem for open-weight model execution: it downloads, orchestrates, and runs open-weight (open-source) AI models on your machine by default; hosted APIs and the Cloud Planner are opt-in. You author models, agents, templates, loops and guardrails in the Console, publish them as versioned configs, and the openeng engine on your machine runs them. You drive it from the Web Console at app.openeng.app.

OpenEng AI

The Console, agents, models, templates, execution flow, cells and configs — the OpenEng AI platform docs.

Quickstart

Sign in, learn the Console, start the engine and publish your first config.

The Console

A tour of every Console page, from Configs and the building blocks to Deploy.

Models

On-device models by default, described by registry, format and runtime, with role defaults.

Cells

Small connectors deployed into your own cloud account, so the engine can reach your cloud without holding your cloud credentials. A known header mismatch blocks Console-registered cells for now.

Govern

Token limits, directories, defaults and the gateway that every hosted call goes through.

The AI engine is published for Linux x86_64 (on Windows, use WSL). It starts with a Console API key and requires an active Pro plan; Pro, Enterprise and pay-as-you-go launch in Q4 2026. See Plans & billing.

Status on September 25, 2026

The OpenEng AI services that the Console and the engine rely on (the OpenEng AI side of the OpenEng API, and the Broker) aren't live yet, so the Console can't load or save your work and no engine can register, whatever the plan. The OpenEng AI pages describe the platform as it works once they are. The engineering apps don't depend on these services. See OpenEng AI.

Where things live

SurfaceWhat you do there
Engineering apps (terminal.openeng.app, k8s.openeng.app, data.openeng.app, git.openeng.app, chat.openeng.app)Use the apps; each drives its own engine on your machine (the Git and Chat engines are coming soon)
Web Console (app.openeng.app)Drive your OpenEng AI engine from the browser
Console (console.openeng.app)Author and publish OpenEng AI configs (tour)
Website (openeng.app)Learn about the platform, and download the Terminal, Kubernetes, Data and AI engines (openeng.app/download)
Billing (billing.openeng.app)Plans and payments (Plans & billing)
Releases (releases.openeng.app)Release notes and version history for the Terminal, Kubernetes and Data engines (Git has a release record but no published binary) (the release feed)
Catalogs (for example models.openeng.app)Browse public agents, skills, templates, models, cells, plugins, loops, library bundles, roles and connectors

Every surface signs in through auth.openeng.app with Google, so one OpenEng account spans all of them.

Browse the docs