Under the hoodTier: how much it should matter to you
3Useful: my rating, 1 to 5
3Signal: worth watching, 1 to 5
ModelsArea: what it touches
Internal ChangesKind: in v2.1.296,
Internal ChangesSection of the release
Unclear Which model this pricing change applies to is not named.
What
One model in Claude Code's model list, the one with a 1-million-token context window and up to 128000 output tokens, now uses a different pricing tier. The new tier, tier_2_10_cache_read_0_10, names its own rate for cache reads. A cache read happens when the model reuses text it has already processed. The old tier was tier_2_10.
Why
The cost Claude Code shows for this model can change, especially in sessions that reuse a lot of cached text.
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 doubtWhich model this pricing change applies to is not named.
