Aion 1.0 Instruct
A small 2.3B Microsoft language model for working right on a device, from the Windows team. It is stated as the foundational small model of the Windows ecosystem for text and operating-system scenarios: fast, private and without a constant connection to the cloud. Languages: English, German, Spanish, Portuguese, Italian, French, Chinese and Japanese, Russian is not on the list. The Hugging Face version is an introduction, and a refined version together with a moderation model will become the Windows system model by the end of January 2027. The full version and an ONNX build for WebGPU are published, and compressed versions are promised later.
- Developer
- Microsoft (Windows Applied Sciences), USA
- First release
- Nov 2026
- Latest release
- Nov 2026
- Sizes
- 2.3B
- License
- Commercial use allowedMIT
- Russian
- Not supported
- Running
- On your own serverAlso runs without a GPU
- Industries
- Customer support, Software development, Documents and accounting
What it does
- an assistant in desktop apps without a cloud
- working with text and system commands offline
- an embedded model for private scenarios
Where it is used
Hardware requirements
Versions
- Aion 1.0 Instruct 2.3B ONNX WebGPU
- Aion 1.0 Instruct 2.3B
How to run it
I can set this up end to end: pick the model size, deploy it on your server and connect it to your systems.
Frequently asked questions
Can Aion 1.0 Instruct be used in a commercial project?
Yes. License: MIT. It allows commercial use, but it is still worth having a lawyer review the license before launch.
What hardware does Aion 1.0 Instruct need?
At minimum: Laptop or regular PC, up to 8 GB of VRAM — smaller versions. Some versions also run on an ordinary CPU, without a GPU. You can calculate the exact VRAM for your model size and context in the hardware calculator.
Does Aion 1.0 Instruct support Russian?
No. The developer documentation lists the supported languages and Russian is not among them.
Where can I download Aion 1.0 Instruct and what does it cost?
The Aion 1.0 Instruct weights are open and free to download. You only pay for the hardware it runs on and for the setup. Source links are at the bottom of this page.
How I deploy it for clients
- SelectionI pick the model size for your task and hardware and test it on your examples.
- DeploymentI deploy it on your server or in a closed network and provide an API.
- Fine-tuningI fine-tune it on your data (LoRA) or connect a knowledge base — whichever is cheaper for the task.
- IntegrationI connect it to your CRM, ERP, bot, website or team chat and set up monitoring.
Similar models
Small Microsoft models trained on carefully selected data: strong at logic and math for their modest size. Versions with images and speech are available.
DetailsTextGemmaGoogle · USACommercial use allowedCompact Google models that run well on a single computer; larger versions understand images. Includes CodeGemma for code, FunctionGemma 270M for function calling and the fast DiffusionGemma.
DetailsTextSmolLMHugging Face · USACommercial use allowedTiny open Hugging Face models for phones and laptops. SmolLM3 (3B) can reason and handle long context; the full training recipe is open.
DetailsSource: huggingface.co/microsoft/Aion-1.0-Instruct-2.3B. Data checked against the model card on 22 Sep 2026. Have a lawyer review the license before commercial launch.


