Follow Discord
Sweep 10 Nov 2026 · 02:18Z Build v2.1.322 539 read Stable v2.1.311 Latest v2.1.322 Next v2.1.322 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.281 ·

Token-usage attachment computes the context window differently

The token_usage attachment now converts the model through an extra step before working out the context window

Under the hood No documentation found Internal Changes
JSON All of v2.1.281
Under the hoodTier: how much it should matter to you
1Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
Context WindowArea: what it touches
Internal ChangesKind: in v2.1.281,
Internal ChangesSection of the release

Unclear The finding does not say what the extra step does or which models it affects.

What

When CLAUDE_CODE_ENABLE_TOKEN_USAGE_ATTACHMENT is set, Claude Code adds a token_usage attachment, which is a note telling the model how much of its context window it has used. The context window is how much text the model can hold at once. Before working out that total, Claude Code now passes the model name through an extra conversion step.

Why

The total in this note may now come out differently for some models.

Read from
How sure we are
Something disagreesSomething we can check disagrees with this entry, or the writer said they could not settle it.
The writer flagged doubtThe finding does not say what the extra step does or which models it affects.

See this entry in the whole of v2.1.281 →

Feedback