GPT-5.6 Sol reaches Amazon Bedrock with a narrower regional footprint

Amazon Web Services says GPT-5.6 Sol, Terra, and Luna can be used through Amazon Bedrock's Responses API. For Sol, AWS lists US East (N. Virginia) and US East (Ohio); Terra and Luna also list US West (Oregon). The practical change is that teams already standardizing on Bedrock can invoke the family through that platform, but Sol is not presented as a region-neutral choice.

OpenAI's model documentation positions Sol as the highest-reasoning tier in the GPT-5.6 family. It accepts text and image input, produces text, has a 1,050,000-token context window, and allows up to 128,000 output tokens. The same documentation lists web search, file search, code interpreter, hosted shell, apply patch, computer use, MCP, and tool search as supported Responses API tools, while audio and video are not supported. That combination makes the Bedrock path most relevant to long-context, tool-driven systems rather than to voice or video workloads.

The comparison with GPT-5.5 is not simply a model-name swap. OpenAI lists Sol at $5 per million input tokens and $30 per million output tokens, and shows the same $5 input price for GPT-5.5. It also warns that prompts above 272,000 input tokens are billed at twice the input rate and 1.5 times the output rate for the whole request. Teams planning to exploit Sol's very large context window should therefore measure their real prompt distribution, cache behavior, and output length before treating the larger limit as a straightforward cost advantage.

AWS says Bedrock pricing matches OpenAI's first-party rates and that usage counts toward existing AWS commitments. Its post also describes zero-operator access, IAM and VPC controls, CloudTrail logging, and a retention period of up to 30 days for classifier-flagged traffic used for automated abuse detection. Those are provider statements rather than an independent compliance assessment, and the announcement does not quantify regional capacity, quotas, or latency. A sensible evaluation is to test the required region, retention terms, tool path, and long-context cost on the team's own workload before changing a production routing policy.