CLOSED, API ONLY · CLAUDE · CLAUDE OPUS · CLAUDE SONNET · CLAUDE HAIKU

Anthropic builds the Claude family of models, offered in tiers such as Opus for the most demanding reasoning, Sonnet for balanced everyday work and Haiku for fast, lower-cost tasks. Claude is known for long-document analysis, careful writing, coding and agentic work, and powers products such as the Claude apps and Claude Code.

Claude is available through Anthropic's own apps and API, Amazon Bedrock, Google Cloud and Microsoft Foundry, so most organisations can reach it through a cloud account they already hold. All Claude models are closed, so there is no self-hosted option, and each access route has its own data handling terms and regional rules.

Anthropic opened a Sydney office and signed a memorandum of understanding with the Australian Government in 2026, and has said local data residency is a frequent request. For now, the verified way to keep Claude processing in Australia is Amazon Bedrock's Australian cross-region inference profiles.

PROCESSING IN AUSTRALIA

Onshore, managed

Most current Claude models via Bedrock's Australian profiles

CHECKED 1 OCTOBER 2026 · CONFIRM BEFORE SENDING SENSITIVE DATA

001 / WAYS TO ACCESS

  1. 01Claude apps (Team and Enterprise)Chat, projects and Claude Code with admin controls, processed on Anthropic's infrastructure.
  2. 02Claude APIDirect developer access with an inference_geo setting limited to US-only or global processing.
  3. 03Amazon BedrockClaude via in-region, Australian geographic (au.) or global cross-region inference profiles.
  4. 04Google Cloud Gemini Enterprise Agent PlatformClaude on Google Cloud (formerly Vertex AI), with no Australian processing listed.
  5. 05Microsoft FoundryClaude hosted on Azure or on Anthropic infrastructure, depending on the model.

KEEPING DATA IN AUSTRALIA

Amazon Bedrock is the verified route for onshore Claude processing. Most current Claude models offer an Australian geographic inference profile (model IDs beginning au.anthropic) that keeps requests from Sydney (ap-southeast-2) or Melbourne (ap-southeast-4) within those two regions for the full inference lifecycle. Some of the newest or specialised Claude models are listed only with global inference from Australian regions, so check each model card before choosing one. Anthropic's own API supports only US or global inference and stores workspace data in the US. Google Cloud's data residency tables do not include Australia for Claude, and Microsoft Foundry offers US data zone controls rather than Australian ones. Confirm by checking the Bedrock model card's Regional availability section for the AU geo profile.

002 / STRENGTHS AND LIMITS

GOOD AT

  • +Strong long-context reading, summarisation and analysis of large documents
  • +Capable coding and agentic performance, including Claude Code for software teams
  • +Clear, controllable writing style suited to professional and customer-facing content
  • +Available across AWS, Google Cloud and Azure as well as Anthropic's own API
  • +Australian geographic inference on Amazon Bedrock for most current models

WATCH OUT FOR

  • –No open-weight versions, so Claude cannot be run on your own servers
  • –Anthropic's own API only offers US or global inference and US workspace storage
  • –Some newest or specialised Claude models on Bedrock are reachable from Australia only via global routing
  • –Google Cloud does not list Claude among the models processed in its Australian regions

BEST FOR

  • Regulated Australian organisations already on AWS that need onshore processing
  • Analysing long contracts, policies, reports and technical documentation
  • Software teams using Claude Code or building coding agents
  • Customer-facing writing where tone and accuracy matter

004 / QUESTIONS

Anything else is quicker to answer on a call. Thirty minutes, free, with the people who would build it.

Yes, through Amazon Bedrock. Calling a supported Claude model with its Australian geographic inference profile from the Sydney or Melbourne region keeps processing within those two regions. Avoid the global profile, which can route requests to any commercial AWS region, and confirm the specific model lists an AU profile.

We choose the deployment that keeps your data where it needs to be, build the app around it and test it before launch.

Book a free call→