Claude Code 2.1.316, released on 31 October 2026, deprecates fast mode for Opus 4.8 as of 30 October 2026. It will be removed on 30 November 2026. After that date, fast mode requests on Opus 4.8 return an error. They do not fall back to standard Opus 4.8. Anthropic’s release note says to migrate to Opus 5.5 to keep the speedup.
Why the detail matters
Most model retirements are quiet: a name changes and nothing breaks. This one is different, because the failure mode is an error, not a slower answer. Any team, script or automated pipeline that has fast mode switched on for Opus 4.8 will simply stop getting responses on 30 November.
Also in the release
The same version adds pooled group budgets to the Claude apps gateway: one monthly budget that a group shares, on top of each person’s own limit. It also adds a CLAUDE_CODE_SUBAGENT_EFFORT setting for how hard sub-agents work by default, and a switch to turn off the ultrathink keyword.
What it means
Pooled budgets and effort settings show how AI development spend is being managed: as a shared, capped resource, not a personal subscription. The deprecation shows the other side. Model choices are now operational dependencies with end dates, and they need an owner.
So what
If a supplier builds your product with Claude Code, ask which models and modes their workflow depends on, and who tracks retirement dates. A 30-day notice is enough to handle calmly, but only if someone is watching. Our AI-assisted development service includes that kind of housekeeping. To talk it through, start a project.