Skip to content

Max plan: Claude Code may compress context at 200K instead of using extended context window #63515

Description

@jimvaras-design

Plan: Max 20x
Model: claude-opus-4-6
Version: 2.1.153

Claude Code triggers context compression ("Transmuting") well before what my Max plan should allow. I suspect the client tracks the context window against the model's base 200K limit and doesn't account for the extended context available on Max plans.

Evidence:

  • Session hits "100% context used" and starts compressing during moderate-length conversations
  • ~82K tokens of system prompt (startup hooks, CLAUDE.md chain, 9 MCP servers) means only ~118K remains for actual conversation before compression
  • On Max 20x, I'd expect access to the full context window, not early compression at 200K

Expected behavior: Claude Code should query the plan's actual context limit from the API and use that as the compression threshold, not a hardcoded model default.

Requested info: Does Max 20x extend the context window for Opus 4.6 beyond 200K? If yes, does Claude Code honor that?

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions