You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
2Signal: worth watching, 1 to 5
Managed SettingsArea: what it touches
ImprovementsKind: in v2.1.293,
ImprovementsSection of the release
Unclear It is not confirmed that the hot-applied mark makes these changes take effect without a restart.
What
The setting that makes the model picker prefer the 1M-context version of a model (modelPrefer1mContext) is now marked as hot-applied for remote policy. Remote policy means settings an organization sends to Claude Code from a server, and hot-applied means a change is meant to take effect while Claude Code is running. The managed setting disableConfigDeprecationWarnings carries the same mark.
Why
When an administrator changes either setting through remote policy, users may no longer need to restart Claude Code for the change to apply.
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.
Confirmed sinceAnthropic's documentation has since written up modelPrefer1mContext, on Deploy Claude Desktop on 3P with an LLM gateway.To show a model as one entry, add `prefer1m: true` next to `supports1m` on its entry. New conversations with that model use the 1M context window. To do this for every model that supports 1M, set [`modelPrefer1mContext`](/docs/third-party/…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 doubtIt is not confirmed that the hot-applied mark makes these changes take effect without a restart.
Anthropic's documentation agreesAnthropic's documentation has since written up modelPrefer1mContext, on Deploy Claude Desktop on 3P with an LLM gateway.
