← 回總覽

Claude 4.6 模型现以标准定价支持 100 万 token 上下文

📅 2026-03-14 13:57 God of Prompt 人工智能 4 分鐘 3939 字 評分: 86
Claude 4.6 Anthropic 上下文窗口 LLM 定价 AI 基础设施
📌 一句话摘要 Anthropic 的 Claude Opus 4.6 和 Sonnet 4.6 已将上下文窗口扩展至 100 万 token,且不收取任何高级附加费。 📝 详细摘要 这条推文强调了 Anthropic 的 Claude 4.6 模型系列的一项重大更新。Opus 和 Sonnet 现在都支持 100 万 token 的上下文窗口,并采用统一费率定价,取消了此前对超过 20 万 token 的输入/输出收取的 2 倍/1.5 倍附加费。此次更新对于大规模代码库分析、长文档审查和扩展 Agent 循环尤其重要。值得注意的是,Opus 4.6 在 100 万 token 规模的
Skip to main content ![Image 1: LogoBestBlogs](https://www.bestblogs.dev/ "BestBlogs.dev")Toggle navigation menu Toggle navigation menuArticlesPodcastsVideosTweetsSourcesNewsletters

⌘K

Change language Switch ThemeSign In

Narrow Mode

Claude 4.6 Models Now Support 1M Token Context at Standard Pricing ==================================================================

Claude 4.6 Models Now Support 1M Token Context at Standard Pricing ================================================================== ![Image 2: God of Prompt](https://www.bestblogs.dev/en/tweets?sourceId=SOURCE_f58696ab) ### God of Prompt

@godofprompt

Claude Opus 4.6 and Sonnet 4.6 now run a full 1 million token context window at standard pricing.

No premium multiplier. No beta header. A 900K-token request costs the same per-token as a 9K one.

What this actually means:

→ Load an entire codebase, hundreds of contracts, or a full agent session into one continuous window

→ Opus 4.6 scores 78.3% on MRCR v2 at 1M tokens, highest recall among frontier models

→ Claude Code users on Max, Team, and Enterprise get 1M by default with 15% fewer compaction events

→ No other model family offers 1M flat-rate pricing across both its top-tier models

Previously, anything past 200K tokens triggered a 2x input / 1.5x output surcharge. That surcharge is gone as of yesterday.

For anyone running long-document review, multi-file code analysis, or extended agent loops, the cost math just changed significantly in Claude’s favor.Show More

!Image 3: Claude

#### Claude

@claudeai · 19h ago

1 million context window: Now generally available for Claude Opus 4.6 and Claude Sonnet 4.6.

!Image 4: Tweet image

1,055

1,724

21.8K

4.1M

Mar 14, 2026, 5:57 AM View on X

6 Replies

2 Retweets

37 Likes

6,591 Views ![Image 5: God of Prompt](https://www.bestblogs.dev/en/tweets?sourceid=f58696ab) God of Prompt @godofprompt

One Sentence Summary

Anthropic's Claude Opus 4.6 and Sonnet 4.6 have expanded their context window to 1 million tokens without premium surcharges.

Summary

This tweet highlights a significant update to Anthropic's Claude 4.6 model family. Both Opus and Sonnet now support a 1-million-token context window at flat-rate pricing, removing the previous 2x/1.5x surcharges for inputs/outputs beyond 200K tokens. The update is particularly impactful for large-scale codebase analysis, long-document reviews, and extended agent loops. Notably, Opus 4.6 achieves a 78.3% recall rate on MRCR v2 at the 1M token scale, setting a high bar for frontier models.

AI Score

86

Influence Score 8

Published At Today

Language

English

Tags

Claude 4.6

Anthropic

Context Window

LLM Pricing

AI Infrastructure HomeArticlesPodcastsVideosTweets

Claude 4.6 Models Now Support 1M Token Context at Standar... ===============

查看原文 → 發佈: 2026-03-14 13:57:44 收錄: 2026-03-14 18:00:48

🤖 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。