{"version":"2.1.307","anchor":"model-context-window-logic-reworked-removing-the-1m-beta-as","canonical_anchor":"model-context-window-logic-reworked-removing-the-1m-beta-as","heading":"Context window size is worked out differently, and [1m] handling was removed","tier":"notice","area":"Context Window","scope":"individual","heads_up":true,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.307\/e\/model-context-window-logic-reworked-removing-the-1m-beta-as","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.307","markdown":"### Context window size is worked out differently, and `[1m]` handling was removed\n\nA model's context window is now the smaller of its stated window and a per-model value, and model names ending in `[1m]` are no longer handled\n\n**Unclear.** Whether a model name ending in `[1m]` still gives a larger window is not settled.\n\n**What**\n\nThe context window is how much text a model can take in at once. Claude Code now sets it to the smaller of two numbers: the window the model states and a window set for each model. A separate value is kept for when you have not opted into a larger window. The old fixed cap was removed.\n\nModel names ending in `[1m]` are no longer handled specially when Claude Code reads model names, follows aliases, or deals with retired models.\n\n**Why**\n\nThis may change how much context you get with a given model and when compaction starts. Compaction is when Claude Code summarises older parts of the conversation to free up room.\n\n- Area: Context Window\n- Tier: You'll notice\n- Useful: 3\/5\n- Signal: 3\/5\n- Scope: individual\n- Heads-up: yes"}