Claude Opus 5.5 is Claude Code's new default, 40% cheaper to run
Anthropic's flagship leads Terminal-Bench at $4/$20 per million tokens, adds five effort tiers and a 2.5x fast mode — and Pro, Max and Team limits just rose.
Copy markdown
The new default Opus — and 40% cheaper to run
Opus 5.5 is now the default Opus in Claude Code (v2.1.280) and ships as `claude-opus-5-5` on the API, AWS, Google Cloud and Azure. It runs 40% cheaper than Opus 5 on typical workloads: $4/$20 per million input/output tokens (down 20% from $5/$25), with cache reads at $0.20 (down 60%). Output is about 30% faster and it uses fewer tokens per task.
Five effort tiers and a 2.5x fast mode
Adaptive thinking now exposes effort levels — low, medium, high, xhigh and max — so you trade reasoning depth for tokens on a per-call basis, and a new fast mode runs up to 2.5x faster for $8/$40 per million. The gains show up on agentic evals: 66.4% on Terminal-Bench 4.0 (up from 52.3) and 57.8% on CursorBench 4.0 (up from 46.6).
Your Pro, Max and Team limits just went up
Alongside the model, Anthropic raised the five-hour usage limits on Pro, Max and Team plans and added a savable rate-limit reset — the small-print change that matters most if you live in Claude Code all day and keep hitting the ceiling.
A court lets the Pentagon flag Anthropic — your keys are fine
A US appeals court upheld the Pentagon's 'supply chain risk' designation of Anthropic, a blow to its federal defense business. It doesn't touch your commercial API, Claude Code or Claude.ai access — this is about government procurement, not your account.