Fable 5.1 lands in Claude Code: longer context, cheaper tokens

Anthropic just released Claude Fable 5.1 and promoted it to the default model in Claude Code. Three changes matter if you write code every day.

Longer context window. Output token allocations are up: long plan write-ups and session recaps no longer get cut mid-flow. On code-audit tasks that read a whole repo, the tail of the report survives.

Chain of thought persisted across tools. When Claude Code chains search_web, then read_file, then edit_file, the reasoning thread is carried from one call to the next instead of restarting. Fewer redundant tool calls, cleaner diffs, faster debugging.

Average −12 % on tokens. On our own test cycles (code editing, refactoring, writing tests), token usage drops by ~12 % at equal task. That compounds with per-plan Anthropic quotas: more usable turns per month.

The Claude Code course was re-run end to end on Fable 5.1 before publication. Its module 18 sums up what you should re-read on your side if you started the course before September 3.