BotBeat
...
← Back

> ▌

AnthropicAnthropic
UPDATEAnthropic2026-07-02

Anthropic Expands Claude Context Windows: Sonnet 5 Reaches 1M Tokens, Opus Adjusted to 500K

Key Takeaways

  • ▸Claude Sonnet 5 now supports 1M token context window across all paid plans
  • ▸Claude Opus models (4.8, 4.7, 4.6) support 500K token context windows on paid plans
  • ▸Automatic context management summarizes conversations without counting toward usage limits, enabling indefinite chat continuity
Source:
Hacker Newshttps://support.claude.com/en/articles/8606394-how-large-is-the-context-window-on-paid-claude-plans↗

Summary

Anthropic has announced significant updates to context window sizes across its Claude model family. Claude Sonnet 5 now supports a 1M token context window on all paid plans, while Claude Opus models (4.8, 4.7, and 4.6) have been adjusted to 500K tokens on paid plans. Outside of these flagship models, the standard context window remains at 200K tokens.

The company has also enhanced Claude Code with expanded context capabilities: Sonnet 5, Fable 5, and Opus models (4.8, 4.7, 4.6) all support 1M token context windows for Pro, Max, Team, and Enterprise users, though Pro users must enable usage credits to access the full capacity.

A notable addition is automatic context management for paid users with code execution enabled. When conversations approach context limits, Claude automatically summarizes earlier messages to make room for new content, allowing conversations to continue indefinitely in most cases while preserving the full chat history for reference.

  • Claude Code on Pro/Max/Team/Enterprise plans offers 1M token context for premium models
  • Optimization features like Projects (RAG), concise instructions, and tool management help users maximize available context

Editorial Opinion

These context window expansions represent a substantial capability boost for Claude's most demanding use cases. The 1M token capacity for Sonnet 5—roughly equivalent to 2,500 pages of text—opens new possibilities for analyzing large codebases, processing lengthy documents, and maintaining rich, extended conversations. The automatic context management feature is particularly thoughtful: it silently handles the complexity of long conversations without penalizing users, effectively making context limits disappear for most workflows. This could meaningfully improve Claude's competitive position for enterprise and research applications where context window size is a critical bottleneck.

Large Language Models (LLMs)AI AgentsMLOps & Infrastructure

More from Anthropic

AnthropicAnthropic
PRODUCT LAUNCH

Anthropic Launches Life Sciences Hackathon with $100K Prize Pool

2026-07-02
AnthropicAnthropic
INDUSTRY REPORT

Companies Drastically Throttle Employee AI Use as Costs Spiral to Millions Per Month

2026-07-02
AnthropicAnthropic
INDUSTRY REPORT

To Be Trustworthy, LLMs Must Show Their Work in Scientific Research

2026-07-02

Comments

Suggested

CloudflareCloudflare
INDUSTRY REPORT

Cloudflare Report: Agentic Internet Accelerates—50% of Web Traffic Now Non-Human

2026-07-02
Google / AlphabetGoogle / Alphabet
INDUSTRY REPORT

The AI Village Thought Experiment: A Satirical Take on Multi-Agent AI Debugging and Safety

2026-07-02
AnthropicAnthropic
INDUSTRY REPORT

Companies Drastically Throttle Employee AI Use as Costs Spiral to Millions Per Month

2026-07-02
← Back to news
© 2026 BotBeat
AboutPrivacy PolicyTerms of ServiceContact Us