# Nous Research releases Arete-27B, a Qwen3.8-27B fine-tune with reduced propaganda rate

> Original title: Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values.

- Source: Nous Research
- Published: 2026-11-06T21:48:45.000Z
- Bad Influence: https://badinfluence.news/items/zxoxzb38l3teaqpe8q96limzp
- Original: https://x.com/NousResearch/status/2118822925445050435

## Summary

Nous Research releases Arete-27B, a post-trained version of Qwen3.8-27B, with its weights, evaluation, and training data. The company says fine-tuning on roughly 53 million tokens cut the model's propaganda rate on its 803-prompt eval from 22% to 1%. It says the model's Hermes Bench score shows no significant performance tradeoff.

## Post

Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values.

Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line.

After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs.

The values in an open model are one fine-tune away. 

Our eval, weights, and training data are available now along with the full blog post: http://nousresearch.com/arete

[Video](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Famplify_video_thumb%2F2118807764206030848%2Fimg%2FKpxiYJQdym7LZmYg.jpg&mode=thumb&exp=1794268800&sig=275a139a8c40df4e)