Microsoft · 2026-06-02 · major
Microsoft Aion 1.0 — On-Device Windows AI Lineup Debuts at Build 2026 With Aion 1.0 Instruct in Edge Insider Today and a 14B Aion 1.0 Plan Reasoning + Tool-Calling Model Shipping In-Box
Microsoft unveiled Aion 1.0 — its on-device Windows AI line — at Build 2026. Aion 1.0 Instruct is a Phi-4-mini-class SLM available in Edge Insider preview today, and Aion 1.0 Plan is a 14B reasoning + tool-calling model shipping in-box on capable Windows PCs.
Microsoft ships a Windows-native AI model line: a small Instruct SLM for everyday text tasks and a 14B agentic Plan model that runs in-box.
Key specs
| Aion 1 plan parameters | 14B |
|---|---|
| Aion 1 plan context window | 32K tokens |
| Aion 1 instruct comparison | smaller and faster than Phi-4-mini |
| Instruct preview channel | Edge Insider |
| Instruct hf target | July 2026 |
| Build 2026 keynote date | 2026-06-02 |
What is it?
Aion 1.0 is Microsoft's new on-device Windows AI model family announced at Build 2026 on June 2. It is distinct from the cloud-hosted MAI lineup also shown the same day. Two variants debuted: Aion 1.0 Instruct, a small language model designed for everyday text intelligence (summarization, rewrite, intents, accessibility), and Aion 1.0 Plan, a 14-billion parameter reasoning and tool-calling model with a 32K context window built for fully agentic workflows running locally.
How does it work?
Aion 1.0 Instruct is smaller and faster than Phi-4-mini and runs across CPUs, GPUs, and NPUs, with a developer preview shipping in the Edge Insider channel and an open-source release planned on Hugging Face in July 2026. Aion 1.0 Plan ships in-box on capable Windows PCs and exposes structured outputs and tool-call orchestration so applications can reason over user intent, invoke tools, manage files, and orchestrate sub-agents. Microsoft pairs the models with an expanded Windows AI APIs surface, including a public-preview Speech Recognition API and Video Super Resolution running on GPU and CPU rather than NPU-only.
Why does it matter?
Until now, on-device LLMs on Windows mostly meant third-party stacks (Llama, Phi via Foundry Local, Liquid LFM). Aion makes a frontier-class agent loop a first-party Windows capability, ships it across the Windows install base, and shores up the local layer of Microsoft's agent push at Build. For app developers it means a Windows-guaranteed agentic runtime they can target without a cloud API key; for users it shifts assistant work — file orchestration, app automation — onto the PC.
Who is it for?
Windows app developers, on-device agent builders, enterprise IT planning local AI deployments
Try it
Edge Insider channels for Aion 1.0 Instruct preview