Anthropic’s Opus 5.5 model is now available on Amazon Bedrock and the Claude Platform on AWS. The release emphasizes cost and efficiency improvements, clearer agentic reasoning for long-running tasks, and built-in safety classifiers.
Opus 5.5 delivers better efficiency than Opus 5, with lower per-token prices and cheaper cache reads that reduce average cost per task.
Opus 5.5 includes new safety classifiers for sensitive domains (biology, cybersecurity, AI development) and will refuse some requests more often than earlier Opus versions.
You can test and call Opus 5.5 via the Amazon Bedrock console (Playground/Test) or programmatically using Bedrock runtime, Converse APIs, AWS CLI/SDKs, or the Anthropic SDK with the Amazon Bedrock token generator.
Anthropic's Claude Opus 5.5 is now available on Amazon Bedrock and Claude Platform on AWS. AWS published guidance on what's new in Opus 5.5, examples for trying the model in the Bedrock console, and code snippets for programmatic access.
Opus 5.5 aims to do more with fewer tokens compared with Opus 5. That efficiency, combined with lower per-token pricing and cheaper cache reads on AWS, reduces the average cost per task. Teams running agentic workflows, longer coding sessions, or extended knowledge-work tasks can run larger or more frequent jobs for the same budget.
Opus 5.5 is positioned as a better fit than Opus 5 for multi-step, long-running sessions where explainability and consistent behavior matter—for example, prolonged software development sessions or long document and report creation.
Opus 5.5 ships with safety classifiers similar to those in Claude Fable 5.1 across three sensitive domains: biology, cybersecurity, and AI development. Because of those classifiers, Opus 5.5 will refuse some requests more frequently than prior Opus models. Teams should expect stricter refusal behavior for content that intersects those domains.
In the Amazon Bedrock console: open Bedrock, choose Test, then Playground, and select Claude Opus 5.5 as the model to run prompts directly.
The blog post provides example Python snippets using boto3 for bedrock-runtime, the converse call, and the Anthropic SDK with the token generator. It also notes Opus 5.5 is a reasoning model whose responses may include a thinking block before the text block, so clients should select the text block rather than assuming a fixed index.
To use Opus 5.5 on AWS you need an active AWS account with Amazon Bedrock access, AWS CLI installed and configured, Python 3.10+, boto3 and the Anthropic SDK installed, and the Amazon Bedrock Token Generator package. Required IAM permissions include bedrock:InvokeModel and bedrock:InvokeModelWithResponseStream.
You can monitor usage and cost through Amazon CloudWatch and AWS Cost Explorer to scale deployments as demand grows.
Claude Opus 5.5 is available on Amazon Bedrock today. Related recent releases on Bedrock include Claude Sonnet 5.5 and Claude Fable 5.1, each targeted at different workload profiles and controls. Snowflake Cortex AI also lists Opus 5.5 in public preview on its platform.
Claude Sonnet 5.5 is now available on Amazon Bedrock and Claude Platform on AWS. It's a smarter, more efficient Sonnet model for focused coding and knowledge work, with a lower cost per task at faster speed. This post covers what's new, when to choose Sonnet, and how to get started.
Claude Fable 5.1 is now available on Amazon Bedrock and Claude Platform on AWS. This post covers Claude Fable 5.1's improvements, the Enterprise Frontier Safeguards for keeping your data in a cloud environment you control, and how to start building with the model on Amazon Bedrock.
Claude Opus 5.5 is in public preview on Snowflake Cortex AI, bringing improvements in coding, financial analysis, and visual understanding.
Claude Opus 5.5 delivers Fable 5.1-level performance at 40% lower cost, with stronger alignment and broader platform availability. See how it stacks up. The post Claude Opus 5.5 Launches: Cheaper, Faster, and Safer Than Ever appeared first on SourceTrail .
Who's become a partner, changed firms, received an award, and more. Share your own announcement!
1. Call for Submissions: Cyber Law Toolkit. The Cyber Law Toolkit, an online resource on international law and cyber activities, is inviting proposals for new scenarios for its next general update, scheduled for September 2027. The Toolkit features a growing collection of hypothetical scenarios inspired by real-world c
Loading more related stories...
Open the app view to save this story, compare related coverage, and continue from the same source.
