← 回總覽

Garry Tan:Opus 4.6 搭配 100 万 Token 上下文实现更复杂的分析

📅 2026-03-21 22:53 Garry Tan 人工智能 2 分鐘 1443 字 評分: 83
Claude Opus 4.6 Claude Code AI 编程 100万 Token
📌 一句话摘要 Garry Tan 强调,搭载 100 万 Token 上下文的 Opus 4.6 大幅缓解了上下文限制问题,让 Claude Code 显著更强大,并表示这是工具链发展的最差时刻,往后只会更好。 📝 详细摘要 这条推文分享了 Garry Tan 关于 Claude Opus 4.6 模型(100 万 Token 上下文窗口)的技术洞察。他回顾去年上下文限制问题曾是持续的痛点,但扩展后的上下文现在能支持完全在上下文内进行更复杂的分析。他还高度评价了 Claude Code 能力的提升,并乐观地表示这代表工具链发展的最低谷,往后只会越来越好。 📊 文章信息 AI 评分:83
![Image 1: Garry Tan](https://www.bestblogs.dev/en/tweets?sourceId=SOURCE_ea220f14)

I underestimated how powerful Opus 4.6 with 1M tokens is. Even last year we were absolutely hitting context limit problems constantly. 1M tokens means you can do much more complex analysis entirely in context. Claude Code is so much better. This is the worst it will ever be.

148 Replies

57 Retweets

1,487 Likes

62.8K Views ![Image 2: Garry Tan](https://www.bestblogs.dev/en/tweets?sourceid=ea220f14)

One Sentence Summary

Garry Tan highlights that Opus 4.6 with 1M tokens context dramatically reduces context limit problems and makes Claude Code significantly better, noting this is the worst it will ever be.

Summary

This tweet shares Garry Tan's technical insight about Claude's Opus 4.6 model with 1 million token context window. He reflects on how context limit problems were constant pain points last year, but the expanded context enables much more complex analysis entirely within context. He also praises Claude Code's improved capabilities and expresses optimism that this represents the lowest point of the tooling's development.

AI Score

86

Influence Score 277

Published At Yesterday

Language

English

Tags

Claude

Opus 4.6

Claude Code

AI Coding

1M Tokens

查看原文 → 發佈: 2026-03-21 22:53:46 收錄: 2026-03-22 02:00:26

🤖 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。