AWS adds xAI’s Grok 4.7 to Amazon Bedrock

AWS has made xAI’s Grok 4.7 available through Amazon Bedrock, giving customers a managed route to the model through the cloud service’s existing runtime interfaces.

The release supports Bedrock’s Responses, Chat Completions, InvokeModel and Converse interfaces. AWS says the Bedrock version accepts text and image inputs and returns text. Requests use the bedrock-runtime endpoint; the bedrock-mantle endpoint is not supported for this model.

Access uses cross-Region inference profiles, which route requests across supported AWS Regions rather than keeping inference within one Region. AWS lists us.xai.grok-4.7 for US geographic routing and global.xai.grok-4.7 for global routing. The AWS model card says in-Region inference is not supported.

AWS and xAI document a 500,000-token context window and four reasoning-effort settings: low, medium, high and xhigh. xAI’s model documentation lists high as the default. xAI calls Grok 4.7 its most capable model for coding and knowledge work and says the model can spend longer on difficult tasks and check its work more carefully. Those performance descriptions are xAI claims, not independent test results.

The Bedrock addition follows xAI’s September 21 Grok 4.7 launch. At launch, xAI said the model was available through Cursor, Grok Build, the Grok API, third-party coding harnesses, model routers and cloud platforms. AWS lists Standard, Priority and Flex service tiers for Grok 4.7 on Bedrock; Reserved capacity is not supported.