xAI's Grok 4.7 lands on Amazon Bedrock with a 500K-token context window

xAI's newest model is now available through the largest enterprise cloud. AWS's machine learning blog announced Grok 4.7 on Amazon Bedrock, and Unite.AI highlighted its 500K-token context window. For enterprises, Bedrock availability matters more than the model card. Grok can now be used under existing AWS contracts, IAM controls and governance tooling, without a separate xAI vendor relationship.

The Bedrock catalog grew quickly in the same period. AWS's September 28 weekly roundup listed:

- OpenAI's GPT-6 Sol, for demanding development and operations work

- GPT-6 Luna, for high-volume repeatable tasks

- Claude Opus 5.5, priced below its predecessor

- the Strands agent harness

- SageMaker HyperPod Inference Gateway, which AWS says cuts first-token latency by up to 82% using GPU-aware routing

- CloudWatch Omni, which adds OpenTelemetry-based AI observability

Bedrock now carries frontier models from OpenAI, Anthropic and xAI, plus Managed Agents built with OpenAI. That makes it the most vendor-neutral of the hyperscaler model platforms.

For xAI, Bedrock is an important enterprise channel. Its Grok 4.6 family was already on the xAI API and GitHub Copilot, with a focus on multi-step coding and the always-on Grok Bot agent. Grok 4.7's 500K context sits between mid-tier models and the 1M-token windows offered by DeepSeek V4.1 Flash and Gemini 4 Argon's output limit.

Caveats: the sources did not include Bedrock pricing or regional availability for Grok 4.7, or independent benchmarks. Enterprise buyers may still weigh Grok's content-moderation history before adopting it, regardless of AWS governance. SpaceX/xAI's membership in NVIDIA's agent-safety coalition may help with that.