← 回總覽

暴论:目前最好的写作模型仍是 Claude Opus/Sonnet 4.6

📅 2026-07-31 01:02 向阳乔木 人工智能 2 分鐘 1424 字 評分: 80
Claude Opus Sonnet 写作模型 模型评测
📌 一句话摘要 作者提出暴论:目前最好的写作模型仍是 Claude Opus/Sonnet 4.6,甚至 Sonnet 更好;Claude Opus 5 存在 Benchmark 刷榜但实际体验一般的问题。 📝 详细摘要 作者提出一个暴论:目前最好的写作模型仍是 Claude Opus/Sonnet 4.6,甚至个人觉得 Sonnet 会更好。大模型都卷 coding 了,比较恶劣的是 Claude Opus 5,Benchmark 刷榜但实际使用体验一般。看来大模型也不是越新就越好。 📊 文章信息 AI 初评:80 来源:向阳乔木(@vista8) 作者:向阳乔木 分类:人工智能 语言
![Image 1: 向阳乔木](https://www.bestblogs.dev/en/tweets?sourceId=SOURCE_50f62a)

@vista8

说个暴论,目前最好的写作模型竟然还是Claude opus/sonnet 4.6 甚至个人觉得 sonnet会更好一点。

大模型都卷coding了,比较恶劣的是 Claude opus 5,Benchmark 刷榜,实际使用体验一般。

看来大模型也不是越新就越好。

64 Replies

3 Retweets

132 Likes

26.7K Views ![Image 2: 向阳乔木](https://www.bestblogs.dev/en/tweets?sourceid=50f62a)

One Sentence Summary

The author puts out a hot take: the best writing model is still Claude Opus/Sonnet 4.6 — even Sonnet might be slightly better; Claude Opus 5 has benchmark gaming issues with mediocre real-world experience.

Summary

The author puts out a hot take: the best writing model is still Claude Opus/Sonnet 4.6 — personally, Sonnet might even be slightly better. All the big models are racing on coding now, and the worst offender is Claude Opus 5, which tops benchmarks but delivers mediocre real-world experience. Looks like newer isn't always better for large models.

AI Screening

80

Influence Score 70

Published Today

Language

Chinese

Tags

Claude

Opus

Sonnet

Writing Model

Model Evaluation

Make your daily reading actually fit you.A daily brief built from the sources you follow. Get started free

查看原文 → 發佈: 2026-07-31 01:02:18 收錄: 2026-07-31 16:00:36

🤖 問 AI

針對這篇文章提問,AI 會根據文章內容回答。按 Ctrl+Enter 送出。