Microsoft Project Zenith Windows developer experience guide

Microsoft formally introduced Project Zenith today, building on the developer-optimized Windows 11 experience it previewed at Build 2026. Zenith is a preconfigured "ready-to-code" setup that ships with coding tools and developer-tuned settings built in. It launches first on AMD's Ryzen AI Halo platform, and AMD showed the first Project Zenith device, a mini PC, on stage at IFA earlier today.

Zenith is tied to "developer-class" hardware with at least 64GB of unified memory and 250+ GB/s of memory bandwidth. Microsoft designed the full Project Zenith experience for new developer-focused devices, with more OEM and silicon partners expected to ship additional systems "in the coming months."

The distinction is important because the underlying software configuration isn't exclusive to Zenith hardware. Microsoft already offers Windows Developer Configurations for any Windows 11 device through WinGet. Project Zenith combines that developer configuration with a defined high-end hardware baseline and an out-of-box setup on purpose-built PCs.

What is Microsoft Project Zenith?

Microsoft describes Zenith as a preconfigured Windows setup for development, paired with tools curated around what developers commonly reach for first. It's a device-and-software experience rather than a new Windows edition or a setting buried somewhere in the Settings app.

Qualifying machines come with preinstalled tools spanning languages and runtimes, source control, and productivity software, including Visual Studio Code, GitHub Copilot, PowerToys, WinApp CLI, and Windows Development Skills, The Verge reports. Windows Terminal and Visual Studio Code are pinned to the Taskbar by default.

Logan Iyer, Microsoft's CVP of Windows Platform and Developer, said the project grew out of direct developer input: "Project Zenith is the result of listening to developers about what they need from Windows today, and where they want the platform to go next."

He also tied the effort to the shift toward AI-assisted coding, adding that Microsoft is "committed to evolving Windows alongside the community that builds on it" as AI continues to transform software development.

Project Zenith Ryzen AI Halo hardware requirements

Qualifying devices need 64GB or more of unified memory and at least 250 GB/s of memory bandwidth, according to the Windows Developer Blog.

AMD's Ryzen AI Halo platform powers the first Project Zenith device, the mini PC AMD revealed at IFA.

Beyond AMD, the lineup is still unnamed. Microsoft says additional OEM and silicon partners will ship Zenith devices using different hardware in the coming months, but it hasn't identified those partners or provided a more specific timeline.

The confirmed baseline so far is:

• Memory: 64GB or more of unified memory

• Bandwidth: 250 GB/s or higher memory bandwidth

• Platform: AMD Ryzen AI Halo first, with other silicon partners coming later

• Form factor: The first announced device is a mini PC

Microsoft and AMD haven't announced pricing, a retail release date, or initial regional availability for the AMD mini PC.

What the Microsoft Project Zenith Windows developer experience includes

Much of the Zenith experience comes from preset configuration rather than entirely new Windows functionality. Microsoft has preconfigured Windows settings across File Explorer, Search, Start, and the Taskbar.

In File Explorer, file extensions, hidden files, the full path in the title bar, and the details pane are visible by default, with long-path support enabled. Recently used files and folders and sync-provider tips are turned off.

Start menu tips and account notifications are disabled too, while PowerToys Command Palette is enabled in Search and Start.

Many of those settings can already be changed manually on ordinary Windows 11 PCs. Microsoft also makes its broader Windows Developer Configuration available separately, allowing developers to install tools and apply developer-oriented settings with a WinGet configuration.

What Zenith changes is the starting point: supported hardware arrives with the tools and configuration already applied.

Local AI development, WSL, and Microsoft's "unmetered" model claims

The more ambitious pitch involves running AI models locally instead of relying entirely on cloud APIs. Microsoft says Zenith's 64GB-plus unified memory and 250 GB/s-plus bandwidth let developers run models with 30 billion or more parameters locally and "unmetered."

In this context, "unmetered" means local inference isn't charged through a cloud provider's per-token billing model. It doesn't mean local compute is free, nor does it establish how fast a particular 30B-plus model will run or how its performance will compare with a cloud-hosted alternative.

Microsoft hasn't published standardized Project Zenith benchmarks alongside that claim, so real-world inference speed, power use, model compatibility, and output performance will depend on the device and workload.

Microsoft also says Zenith devices benefit from its existing platform investments for developing and running AI agents.

Windows Subsystem for Linux is part of that foundation. Microsoft open-sourced WSL in 2025 and has since expanded its container tooling so developers can build, run, and interact with Linux containers directly through WSL.

That container support isn't generally available yet, however. Microsoft's current WSL container documentation says the feature requires WSL 2.9.3 or later and is still available only through a pre-release build.

WSL containers also predate Project Zenith. They are part of Microsoft's broader Windows developer platform rather than a capability exclusive to Zenith hardware.

What's still missing from Project Zenith

Pricing, retail availability, and details about the additional OEM and silicon-partner devices are still missing.

Microsoft also hasn't explained exactly how much Project Zenith hardware will vary between vendors or whether every device will ship with an identical tool set beyond the baseline experience it announced.

The local-AI claim carries another unanswered question. Running a 30B-plus parameter model locally without cloud token charges is one of Zenith's headline advantages, but Microsoft and AMD haven't published standardized third-party-comparable benchmarks showing inference speed or performance across real developer workloads.

Project Zenith therefore establishes a hardware floor and a preconfigured Windows development environment, but not yet a complete buying picture.

The first systems will use AMD Ryzen AI Halo, with other devices promised in the coming months. Pricing, availability, additional hardware partners, and independent performance testing will determine how useful that combination proves outside Microsoft's launch demonstrations.