xAI's Grok 4.6 lands on Amazon Bedrock with 500K context and new xhigh reasoning tier
AWS ML Blog · rss · 2026-09-22
AWS announced that xAI's Grok 4.6 is now available in Amazon Bedrock (launched August 18, 2026), xAI's second model there, offering a 500K token context window and configurable reasoning effort across four levels: low, medium, high, and xhigh.
- Model focus: Builds on Grok 4.5 for long-running agents and interactive/visual work. xAI reports the model shows more self-testing and verification on long trajectories and stronger first passes on visual projects. Training involved a longer supplemental run, SFT trajectories regenerated with Grok 4.5, and agentic RL tasks spanning coding, kernel optimization, and web development.
- New on Bedrock: Served on both bedrock-mantle and bedrock-runtime endpoints, with Converse API including streaming; cross-region inference profiles us.xai.grok-4.6 ($2.20/M input) and global.xai.grok-4.6 ($2.00/M, cheaper default). Also supports Bedrock Guardrails, invocation logging to CloudWatch (including reasoning tokens), and prompt caching at roughly a quarter of standard input price.
- Benchmarks: Reported Grok 4.6 High scores include AA Intelligence Index 61, CursorBench v3.2 69.9%, Terminal-Bench v3.0 26%.
- Limits: Text and image input, text output only; no audio, video, image generation, or embeddings.
More from Infra
- OpenAI removes Ultrafast tier from GPT-5.6 Sol in Codex, fueling GPT-6 Sol rumors — imjustnewatai · 2026-09-22
- Is upgrading from 2x to 4x RTX 3090 worth it for local LLM work? — fgoricha · 2026-09-22
- Qwen-Image local on a 24GB MacBook Pro takes 5-6 minutes per image — vista8 · 2026-09-22
- Cerebras CEO on Jensen Huang: a decade trading as 'nobody' before Nvidia made it — rohanpaul_ai · 2026-09-22
- How Tencent Hunyuan packed a 770B model into 214 GiB with 5-bit-per-4-weights quantization — TencentHunyuan · 2026-09-22
- 456GB DeepSeek v4.1 runs locally at 40 tok/s with Threadripper + dual RTX 6000 hybrid setup — HankYeomans · 2026-09-22