AWS says new OpenAI and Anthropic models have landed on Bedrock
AWS has used a weekly roundup to outline claimed AI model additions and cloud service updates, including new Bedrock models, agent observability and EventBridge changes.
Amazon Web Services has claimed a set of generative AI and cloud operations updates in a weekly roundup published on its AWS News Blog, led by new foundation models on Amazon Bedrock.
According to the company post, Bedrock has added GPT-6 Sol and GPT-6 Luna from OpenAI, alongside Claude Opus 5.5 from Anthropic. AWS described GPT-6 Sol as aimed at demanding, recurring development and operations work, while GPT-6 Luna is positioned for focused, repeatable tasks at high volume. The blog also claimed both OpenAI models ship at significantly lower pricing than their GPT-5.6 predecessors, although the supplied text does not provide figures.
AWS said Claude Opus 5.5 is the first model in Anthropic’s Claude 5.5 family and is tuned for agentic coding and long-running tasks. It also claimed the model does more with fewer tokens than Opus 5. The supplied material does not include independent benchmarks or third-party verification of those performance claims.
The roundup also described Amazon CloudWatch Omni, which AWS said provides a single collaborative observability experience for applications and AI agents. The service is said to use OpenTelemetry, support enterprise single sign-on without console access, auto-discover services and bring AWS DevOps Agent into investigations.
Other updates included enhanced custom event buses in Amazon EventBridge, with features such as a centralised bus across accounts, optional event ordering, content-based deduplication and synchronous invocation for targets including AWS Lambda. AWS also outlined the SageMaker HyperPod Inference Gateway, described as a Kubernetes-native, GPU-aware routing layer for large language model inference.
The post additionally said AWS End User Messaging and Amazon SES now publish AI agent skills for AWS MCP Server, allowing supported coding agents to guide messaging workflows. The supplied extract begins to mention a Strands harness release, but does not include enough detail to assess that update.
Comments 0
Be the first to comment.