Overview
Claude Code version 2.1.75 expanded the default context window to 1 million tokens for Max, Team, and Enterprise plan subscribers. The change shipped in the 2.1.75 changelog entry and applies automatically without any configuration change on the user's end.
Previously, the 1M context window for Opus 4.6 required extra usage allowance beyond the standard plan limits. This release reclassifies it as the default, meaning all eligible plan tiers now get the full window by default. This is a capacity change tied to a specific model and plan combination, not a general API upgrade.
What this means for brands
A 1M token context window means Claude Code can now process entire large codebases, lengthy documentation sets, or multi-file brand style guides in a single session without truncation. For brands that appear in developer-facing content, technical documentation, or SDK references, this increases the likelihood that Claude surfaces brand information that previously fell outside the active context. Operators running brand audits or competitive analysis via Claude Code can now include substantially more source material in a single query, which changes how comprehensively Claude can reason about brand mentions across a corpus.
The practical ceiling shift also affects how Claude handles long retrieval-augmented generation tasks. If your brand is documented in large technical specs, API references, or multi-chapter guides, those documents are now more likely to be fully ingested rather than chunked or summarized, which affects the accuracy and completeness of any brand-related outputs Claude produces from that material.
What to do
If you use Claude Code for brand monitoring or content audits, test whether your full documentation corpus now fits within a single context window. Load your complete brand guidelines, product documentation, or style guide into a Claude Code session on a Max, Team, or Enterprise account and confirm the token count stays under 1M. If it does, you can retire any chunking logic you built to work around the previous limit. Re-run any brand query benchmarks you have from before 2.1.75 to check whether longer context changes how Claude describes or positions your brand when given more material to work with. If you manage a developer-tools brand with presence in large open-source codebases, run a few representative queries against the expanded window to see whether citation patterns or brand placement in responses shifts now that Claude can hold more code context at once.