# Qwen 3.8 Flash outperforms expectations on simple accuracy tests

> Original title: 在一些简单的准确性任务评测里，Qwen 3.8 Flash这个模型意外的强...

- Source: Khazix
- Published: 2026-10-30T06:28:40.000Z
- Bad Influence: https://badinfluence.news/items/ephm1bdhg6aj2rta6euf1cy7z
- Original: https://x.com/Khazix0918/status/2116054660851311061

## Summary

Khazix (@Khazix0918) says Qwen 3.8 Flash performed unexpectedly well on some simple accuracy evaluations. He reports that many of his team's product automation tasks have shifted from Claude Opus 5.5 to Qwen 3.8 Flash, with more than half now switched.

## Post

在一些简单的准确性任务评测里，Qwen 3.8 Flash这个模型意外的强... 
今天看了一下，我们现在各种产品的很多自动化任务，居然已经一大半都被Claude Opus 5.5切换成Qwen 3.8 Flash了...

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHV28cqsbgAAH0-N.jpg&mode=thumb&exp=1793577600&sig=529a904062f4df8d)