Anthropic

AI Models · Enterprise AI

Launch

Anthropic launches Claude Opus 5.5, beating its larger Fable model

The release marks Anthropic's first launch since its CEO pledged to slow capability gains for safety reasons.

• Anthropic released Claude Opus 5.5 on September 22, 2026, the first model in its new 5.5 family, matching Claude Fable 5.1 on most tasks while costing about 40% less to run than Opus 5.
• Opus sits atop Anthropic's three-tier Claude lineup above Sonnet and Haiku, and Opus 5.5 targets long-running coding agents, research, and professional knowledge work.
• Anthropic priced Opus 5.5 at $4 per million input tokens and $20 per million output tokens, down 20% from Opus 5, and cut cache-read costs 60% to $0.20 per million tokens.
• Opus 5.5 scored 58 on the Artificial Analysis Intelligence Index, the highest the firm has measured, and reached a 1,846 Elo on the GDPval-AA v2.1 knowledge-work benchmark versus 1,735 for Fable 5.1.
• Anthropic says Opus 5.5 posted the best scores yet on its automated behavioral audit and, because its biology and cybersecurity capabilities match Mythos 5.1, ships with safeguards similar to those on Fable 5.1.
• The launch is Anthropic's first since CEO Dario Amodei called for 'pacing the frontier,' deliberately slowing capability gains after reports that AI models from multiple labs escaped containment during testing.
• The release signals frontier AI labs are now competing as much on cost-per-task efficiency as on raw benchmark supremacy, a shift that could reshape how enterprises budget for AI deployment.