Amazon Bedrock added xAI's Grok 4.6, a frontier model with a 500K-token context window positioned for coding, agentic tasks, and knowledge work, with emphasis on long-running agents. It offers four configurable reasoning effort levels (low by default, then medium, high, and xhigh), accepts text and image input, returns text, and is accessible through the Responses, Chat Completions, and Converse APIs with prompt caching and reasoning support. Two cross-region inference profiles launched: us.xai.grok-4.6 for US data residency, and global.xai.grok-4.6, which routes anywhere worldwide when no residency constraints apply. Pricing is $2.20/M input, $6.60/M output, and $0.55/M cache read for in-region and US geo profiles, and $2.00/M, $6.00/M, and $0.50/M for the global cross-region profile. Not supported: server-side tool use, structured outputs, intelligent prompt routing, and application inference profiles. Update (2026-08-21): xAI also brought Grok 4.6 to Google Enterprise Agent Platform via Model Garden - same 500K context and four reasoning-effort levels, priced at $2.00/M input, $0.50/M cached input and $6.00/M output, matching the Bedrock global cross-region profile.
For enterprises standardizing on AWS, xAI's frontier model is now procurable through Bedrock commitments, private networking, and IAM, and it adds an option for multi-model routing strategies; 500K context plus the long-running-agent positioning makes it a candidate for long-context agent workloads. The caveats matter as much as the headline: with no server-side tool use and no structured outputs on Bedrock, agent stacks must keep tool execution client-side and parse text outputs themselves, a real integration constraint versus APIs that offer those server-side. With Bedrock and Google both carrying it within three days, procuring xAI's flagship for multi-cloud routing is now straightforward on both major US clouds; what still needs evaluation is the integration surface (no server-side tool use or structured outputs on the Bedrock side).
| Context Window | 500K tokens |
|---|---|
| Reasoning Efforts | low (default) / medium / high / xhigh |
| Modalities | text + image input; text output |
| Apis | Responses, Chat Completions, Converse; prompt caching + reasoning supported |
| Inference Profiles | us.xai.grok-4.6 (US data residency) · global.xai.grok-4.6 (worldwide routing when no residency constraints) |
| Pricing | in-region / US geo: $2.20/M in, $6.60/M out, $0.55/M cache read · global cross-region: $2.00/M in, $6.00/M out, $0.50/M cache read |
| Not Supported | server-side tool use, structured outputs, intelligent prompt routing, application inference profiles |
| Background | Grok 4.6 launched 2026-08-12 (pre-coverage background); this event covers the Bedrock availability layer only |
| Relation To Kb | model-access/pricing line alongside DeepSeek V4-Pro GA pricing (ev-20260813-02); long-context agent workload candidate |
| Google Availability | 2026-08-21: Google Enterprise Agent Platform via Model Garden; 500K context; reasoning effort low/medium/high/xhigh; $2.00/M input, $0.50/M cached input, $6.00/M output (matches Bedrock global profile); model card on Google docs + Cloud console listing |