You'll noticeTier: how much it should matter to you
4Useful: my rating, 1 to 5
5Signal: worth watching, 1 to 5
ModelsArea: what it touches
ImprovementsKind: in v2.1.282,
What probably matters to youSection of the release
Unclear Which screen reads this built-in list, and whether a copy from the server overrides it, is not settled.
What
The model list built into Claude Code, and the places that show model names, changed in several ways:
claude-opus-5-5 ("Opus 5.5") is now described as "Most capable for ambitious work" instead of "For complex tasks".
claude-opus-5 moves from section: "main" to section: "overflow", loses its "For complex tasks" description, and loses quick_select. This is the entry with the Fast mode toggle, 1M input tokens and 128000 output tokens. claude-opus-5-5 and claude-sonnet-5 keep quick_select.
The "Opus" and "Opus (1M context)" entries in the model picker no longer hard-code "Opus 5.5" or claude-opus-5-5. They show the id and display name of whatever the opus alias currently points to. The checks for Sonnet and Opus also compare against the aliases rather than the literal claude-sonnet-5 and claude-opus-5-5 ids.
The example in the Schedule (cloud routines) skill now fills "model" with whatever the sonnet alias resolves to, instead of claude-sonnet-5.
Why
Opus 5 may no longer appear among the quick picks in /model and may sit in a secondary group. Picker labels and routines built from the example track the current aliases, so they change when an alias moves to a newer model.
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Added sinceA small documentation edit on prompting-claude-opus-5-5 touched a line naming claude-opus-5 after this was published.* **Reasoning extraction:** Requests that push the model to reproduce its internal reasoning in the response text may be declined with the `reasoning_extraction` category. If your prompts ask the model to write out its reasoning in the res…build-with-claude/prompt-engineering/prompting-claude-opus-5-5see the edit
Confirmed sinceAnthropic's documentation has since written up claude-opus-5, on Deploy Claude Desktop on 3P with an LLM gateway.When `inferenceModels` is unset, Claude Desktop on 3P populates the model picker from your gateway's `GET /v1/models` response. Auto-discovery shows only models whose IDs are recognizably Claude. Set [`inferenceModels`](/docs/third-party/c…third-party/claude-desktop/gatewaysee the edit
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 screen reads this built-in list, and whether a copy from the server overrides it, is not settled.
Anthropic's documentation agreesAnthropic's documentation has since written up claude-opus-5, on Deploy Claude Desktop on 3P with an LLM gateway.