OpenAI GPT-5.6 arrives on Amazon Bedrock with cross-Region inference in 25+ Regions
AWS ML Blog · rss · 2026-08-21
OpenAI GPT-5.6 models are now available on Amazon Bedrock in 25+ AWS Regions with cross-Region inference (CRIS), per an AWS ML Blog post co-written with OpenAI. Three general-purpose variants — Sol, Terra, and Luna — are tuned for different capability/cost balances, all supporting text+image input, a 1M token context window, reasoning mode, server-side tool calling, and prompt caching.
CRIS works via inference profiles: a geographic profile (e.g., us.openai.gpt-5.6-terra) routes only within one geography for data residency; a global profile (global. prefix) routes to any supported commercial Region based on real-time capacity for the widest compute pool. Billing and quota are tracked at the account level.
Developers can point an existing OpenAI SDK client at Bedrock's OpenAI-compatible endpoint and swap the model parameter for the profile ID, using the Responses API, Chat Completions API, or Bedrock Converse API. For production, the aws-bedrock-token-generator package derives short-term bearer tokens from AWS credentials.
More from Infra
- Haight Street Compute Market: a one-day trading floor for GPU deals in SF — ctjlewis · 2026-10-16
- Temporal Serverless Workers on Cloud Run: queue-driven autoscaling, scale to zero — rseroter · 2026-10-16
- LMA brings day-0 model support to TPU for the first time, spanning NVIDIA, AMD, Intel via SGLang — ying11231 · 2026-10-16
- Semtech's analyst-day slide cleanly explains the four signal-integrity impairments — BenBajarin · 2026-10-16
- Fortytwo Engine boosts Qwen3.8-27B accuracy 7% and cuts tokens 25% via Swarm Inference — rohanpaul_ai · 2026-10-16
- Berkeley scholars: concentrated chip supply chain makes a hardwired frontier AI pause feasible — wfithian · 2026-10-16