{"version":"2.1.321","anchor":"model-usage-records-carry-contextwindow-canonicalmodel-pro","canonical_anchor":"model-usage-records-carry-contextwindow-canonicalmodel-pro","heading":"Cost tracking records more details about each model","tier":"internal","area":"Elsewhere","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.321\/e\/model-usage-records-carry-contextwindow-canonicalmodel-pro","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.321","markdown":"### Cost tracking records more details about each model\n\nClaude Code's cost tracking now records each model's context window, output limit, standard name, provider and cost basis\n\n**What**\n\nWhen Claude Code tracks the cost of using each model, it now passes along more details about the model:\n\n- context window, how much text the model can take in at once\n\n- maximum output tokens, how much it can write in one reply\n\n- its standard model name\n\n- the provider\n\n- the cost basis\n\nThe counts it sends to OpenTelemetry, a standard way of sending usage figures to monitoring tools, are now sent slightly later rather than immediately.\n\n**Why**\n\nPer-model cost records carry more detail than before.\n\n- Area: Elsewhere\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}