Inference Services

AI inference APIs and services available to Australian users: latency from Australia, data residency guarantees, and models offered.

Total Services

8

tracked services
AU-Hosted

4

with AU data residency↑ up
Avg AU Latency

15 ms

AU-hosted services→ stable
Offshore

4

US-routed services→ stable

Service Directory

Showing 8 of 8
AWS Bedrock (Sydney)
AU HostedOffshore Owned
Amazon Web Services
AU latency: 12ms
3 models available
Azure OpenAI (Australia East)
AU HostedOffshore Owned
Microsoft Azure
AU latency: 15ms
3 models available
Google Vertex AI (Sydney)
AU HostedOffshore Owned
Google Cloud
AU latency: 14ms
3 models available
OpenAI API (US-hosted)
Offshore Owned
OpenAI
AU latency: 185ms
4 models available
Anthropic API (US-hosted)
Offshore Owned
Anthropic
AU latency: 190ms
3 models available
Oracle AI Services (Sydney)
AU HostedOffshore Owned
Oracle Cloud
AU latency: 18ms
3 models available
Groq Cloud (US-hosted)
Offshore Owned
Groq
AU latency: 195ms
2 models available
Together AI (US-hosted)
Offshore Owned
Together AI
AU latency: 200ms
3 models available

Latency from Australia (ms) — AU hosted vs offshore

Latency Impact

AU-hosted services average ~14 ms round-trip from Australian clients. Offshore US services average ~185 ms — a 13x latency penalty.

For real-time applications (voice AI, code completion), latency >50 ms degrades user experience significantly.

Latency comparison

AWS Bedrock (Sydney)
12 ms
Azure OpenAI (Australia East)
15 ms
Google Vertex AI (Sydney)
14 ms
OpenAI API (US-hosted)
185 ms
Anthropic API (US-hosted)
190 ms
Oracle AI Services (Sydney)
18 ms
Groq Cloud (US-hosted)
195 ms
Together AI (US-hosted)
200 ms