Grok 4.7 Is Now Available on Amazon Bedrock

AWS announced on September 28 that xAI's Grok 4.7 is generally available on Amazon Bedrock, one week after xAI launched the model on its own API. The AWS machine learning blog and model card lay out the integration details, and the model is also listed in Bedrock's documentation alongside the other xAI models the service hosts.

The technical shape: Grok 4.7 accepts text and image input and returns text, with a 500K-token context window and four configurable reasoning-effort levels from low up to xhigh, letting developers trade latency and cost against reasoning depth. On Bedrock the model is served on the bedrock-runtime endpoint through cross-region inference profiles, so requests must name us.xai.grok-4.7 or global.xai.grok-4.7 rather than a bare model ID. It supports the Responses, Chat Completions, InvokeModel, and Converse APIs.

AWS also relayed a safety claim from xAI: Grok 4.7 was built with a new safeguard stack and, by xAI's own testing, is its strongest model on refusals and resistance to jailbreaks, framed around staying useful for legitimate cybersecurity and biology work while refusing dangerous requests. That is the vendor's characterization, published through AWS rather than independently evaluated.

What is demonstrated versus claimed: availability is real and verifiable in the model card, but benchmark superlatives and pricing come from xAI. Its launch coverage lists the model at $2 per million input tokens and $6 per million output tokens, the same rates as Grok 4.6, though Bedrock region availability and service-tier pricing should be confirmed in the AWS console before committing workloads.

Why it matters: distribution is becoming the real battleground for frontier models. Enterprises that standardize on AWS get Grok under IAM, VPC, logging, and billing controls they already run, no separate xAI contract required. Every frontier lab is now racing to be one click away inside the big three clouds, and Bedrock's model menu is quietly becoming the default procurement catalog for enterprise AI.