Opus 5.5: A Major Step Forward in Intelligence, Efficiency, and Value
We’ve just launched Opus 5.5, a major improvement for our customers. It sets a new standard for useful intelligence, delivers significant cost savings, and earned the highest alignment scores on our most comprehensive test to date.
Built for Real-World Work
Enterprise customers are already seeing meaningful improvements across a wide range of use cases, including:
- Research
- Agentic coding
- Financial analysis
- Complex reasoning
- Organizational workflows
Opus 5.5 is diligent, thorough, and clearer in how it communicates, helping teams move from ideas to reliable outcomes with greater confidence.
More Efficient, More Affordable
Opus 5.5 is also significantly more efficient. It uses less compute to serve than Opus 5, and we’re passing those savings on through our pricing.
Combined with improved token efficiency, typical workloads are expected to cost approximately 40% less than Opus 5.
We’re also increasing five-hour usage limits for customers on Pro, Max, and Team plans.
Designed for Enterprise Reliability
Opus 5.5 reflects our continued focus on delivering strong value for enterprises using AI in real-world environments.
It combines improved capability and efficiency with the safety and reliability organizations need when deploying AI across their teams and operations.
And this is only the beginning. We’ll have more to share soon about Sonnet 5.5 and Haiku 5.5.
Stay tuned.