Claude Opus 5.5 (Anthropic)

Anthropic has released Claude Opus 5.5, the first model in its new 5.5 generation, on September 22, 2026, just two months after Opus 5. The company describes it as its most powerful model yet, with major gains in coding and agent-based tasks while costing roughly 40 percent less than its predecessor. Pricing now stands at $4 per million input tokens and $20 per million output tokens, with cache reads reduced to $0.20 per million. Anthropic reports text generation over 30 percent faster than Opus 5, benchmark gains on Terminal Bench 4.0, FrontierCode v1.1, and GDPval-AA v2.1, and 85 percent fewer attempts to circumvent security boundaries. Sonnet 5.5 and Haiku 5.5 are expected in the coming weeks.

Claude Opus 5.5, Anthropic's latest Opus model, is now available in public preview on Snowflake Cortex AI with same-day availability across CoCo, Cortex Agents, AI Functions, Cortex Inference, and CoWork. All inference calls remain within Snowflake's security and governance perimeter, allowing enterprise teams to use t According to Anthropic's reported figures, Opus 5.5 shows improvements over Opus 5 in complex coding and knowledge work: one tester completed a 680,000-line code migration in under a day, and another audited a 200,000-line codebase in under three hours (versus over 20 hours and 2.5x more tokens for Opus 5). It scores 6

Anthropic announced Claude Opus 5.5 on September 22, 2026, framing it as its new flagship Opus model and the first release in the Claude 5.5 family. The article positions the launch as a notable shift because it pairs higher capability with reduced operating cost, making repeated use inside AI agents more practical. According to the article, Opus 5.5 is aimed at coding, professional work, computer use, and long-running agents, with Anthropic claiming it matches Claude Fable 5.1 on most tasks while costing 40% less than Opus 5 on typical workloads. Access is offered through Claude, Anthropic's developer platform, AWS, Google Cloud, and Microsoft Foundry.

Anthropic launched Claude Opus 5.5 on September 22, 2026, positioning it as its flagship model for long-running agentic coding and complex knowledge work. The company says the model costs 40% less to run than Opus 5 on typical workloads and was tested against GitHub Copilot CLI and Visual Studio Code environments. Available through the Claude API under the identifier claude-opus-5-5, Opus 5.5 lists pricing at $4 and $5 per million tokens, roughly 20% below Opus 5's listed rates, and now supports five effort levels defaulting to medium instead of high. Anthropic recommends starting with Opus 5.5 for most workloads but has not retired Opus 5, Opus 4.8, or Opus 4.7, and independent production benchmarks remain limited at launch.

Claude Opus 5.5 takes the top spot on the Artificial Analysis Intelligence Index with a score of 58 at maximum effort, the highest the firm has measured and several points ahead of the next model. It leads six of ten index evaluations, including Humanity's Last Exam at 61.4% (up from Fable 5.1's 59.1%) and SciCode at 6 Anthropic has cut Opus 5.5 prices substantially, with input at $4 per million tokens and output at $20 per million, a 20% reduction versus Opus 5, while cache read discounts are steeper still and account for most of the cost in coding and agent workloads. The release comes just days after CEO Dario Amodei's essay calli