The AI arena is free today

Open Superagent
MicrosoftReleased on Nov 18, 2026

Aion 1.0 Instruct: Benchmarks, Pricing & Context Window

Aion 1.0 Instruct is a language model from Microsoft, released in November 2026.

Aion 1.0 Instruct is a 2.3B-parameter open-weights small language model from the Microsoft Windows Applied Sciences Group, published on Hugging Face on 2026-11-18 as microsoft/Aion-1.0-Instruct-2.3B under the MIT license. It is the

Aion 1.0 Instruct benchmarks

Capability tiers

Standing within each category, adjusted for leaderboard depth.

Real tasks performance

High-confidence performance for Aion 1.0 Instruct across real-world prompt categories. Only 95% intervals at most 4 points wide are shown.

Performance by conversation depth

How Aion 1.0 Instruct holds up as conversations get longer.

Quality Tracker

Aion 1.0 Instruct Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

LLM Stats Logollm-stats.com - Thu Nov 19 2026
Notice missing or incorrect data?

Aion 1.0 Instruct model size

Aion 1.0 Instruct has 2.3 billion parameters. See how it compares to other models in the same parameter range.

Parameters
2.3B
Small (<3B)
2.3B
1B7B70B405B

Try now

huggle
Aion 1.0 Instructin Huggle

Make it with
Aion 1.0 Instruct.

Aion 1.0 Instruct

Aion 1.0 Instruct latency

Aion 1.0 Instruct time to first token, sustained output throughput, and failed-request rate from live model usage over the trailing 7 days.

Aion 1.0 Instruct examples

Recent arena outputs from Aion 1.0 Instruct, picked from the highest-ranked matchups.

Aion 1.0 Instruct license

Aion 1.0 Instruct is released under the MIT license, which permits commercial use, has 2.3B parameters, has a knowledge cutoff of December 2025.

License
MIT
Commercial use allowed
Parameters
2.3B
Knowledge cutoff
December 2025

MIT License - allows commercial use

Aion 1.0 Instruct resources

Official sources for Aion 1.0 Instruct: official launch post, source repository.

Aion 1.0 Instruct vs other models

The most-compared alternatives to Aion 1.0 Instruct are QwQ-32B, Qwen3 VL 8B Instruct, Qwen2.5 72B Instruct. Open any pair side-by-side for benchmarks, pricing, context, and latency.

Models like Aion 1.0 Instruct

Models ranked just above and below Aion 1.0 Instruct by LLM Stats score.

 

QwQ-32B

Score pending
 

Qwen3 VL 8B Instruct

Score pending
 

Qwen2.5 72B Instruct

Score pending
 

Llama 3.1 8B Instruct

Score pending
 

Gemini 1.0 Pro

Score pending
 

Phi-3.5-mini-instruct

Score pending

FAQ

Common questions about Aion 1.0 Instruct.

When was Aion 1.0 Instruct released?

Aion 1.0 Instruct was released on November 18, 2026 by Microsoft. This is the official Aion 1.0 Instruct release date tracked on LLM Stats.

How big is Aion 1.0 Instruct?

Aion 1.0 Instruct has 2.3 billion parameters. It ships as an open-weight model, so you can download and run it on your own hardware.

Who created Aion 1.0 Instruct?

Aion 1.0 Instruct was created by Microsoft.

What is the license for Aion 1.0 Instruct?

Aion 1.0 Instruct is released under the MIT license. This is an open-source / open-weight license that permits self-hosting.

What is the knowledge cutoff date for Aion 1.0 Instruct?

Aion 1.0 Instruct has a knowledge cutoff of December 2025, meaning it was trained on data up to that point and may not know about events after it.

Where is the Aion 1.0 Instruct paper or technical report?

Aion 1.0 Instruct has a paper or technical report available at https://huggingface.co/microsoft/Aion-1.0-Instruct-2.3B. Use that source for architecture, training, release and evaluation details.

What models should I compare Aion 1.0 Instruct against?

Common Aion 1.0 Instruct comparisons include Aion 1.0 Instruct vs QwQ-32B, Aion 1.0 Instruct vs Qwen3 VL 8B Instruct, Aion 1.0 Instruct vs Qwen2.5 72B Instruct. Compare them side by side for benchmark scores, pricing, context window, latency and provider availability.