Claude Opus 5.5: frontier performance at lower cost

Opus 5.5 reports 66.4% on Terminal-Bench 4.0, 81.8% on OSWorld 2.1 and matches Fable 5.1 on most work for ~40% less than Opus 5.

What happened and why it matters

Notably, Anthropic cautioned that benchmark margins are an increasingly unreliable guide to real-world differences — itself a sign that benchmarks are saturating faster than they can be built.

Sources

• Introducing Claude Opus 5.5 ↗ Anthropic · 22 Sep 2026