{"version":"2.1.281","anchor":"context-window-resolution-refactored-to-a-model-descriptor","canonical_anchor":"context-window-resolution-refactored-to-a-model-descriptor","heading":"Context-window resolution refactored to a model descriptor","tier":"internal","area":"Compaction","url":"https:\/\/changelogs.core-directive.com\/v\/2.1.281\/e\/context-window-resolution-refactored-to-a-model-descriptor","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.281","markdown":"### Context-window resolution refactored to a model descriptor\n\nThe code that picks the context window size for auto-compact now works from a full model description instead of a model name\n\n**Unclear.** The finding does not settle whether the new descriptor fields change the window size chosen for any model.\n\n**What**\n\nThe context window is how much conversation the model can take in at once. Auto-compact shortens the conversation automatically when it gets close to that limit. The code that decides the window size for auto-compact now takes a description of the model instead of only its name. The description includes the model's window size, whether it has a native 1M-token window, whether a 1M opt-in is honored, whether usage past 200k tokens is billed, and whether the model is recognized. The labels showing where the size came from (env, settings, clientdata, experiment, model-default, unknown-model, auto) are unchanged.\n\n**Why**\n\nThis is internal restructuring. The finding does not describe any change in when auto-compact happens.\n\n- Area: Compaction\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5"}