The latest open-weight models
Frontier and open-weight catalogue, hosted in ap-syd-2 and ap-mel-1.
Coming Soon
Sovereign multimodel inferencing.
Run frontier and open-weight models on Australian-controlled infrastructure. Sovereign Australian inference, AUD pricing, zero retention. No prompt leaves the country.
What it is One API.
One API.
One hundred sovereign models.
A single Australian-hosted endpoint that serves Llama, Mistral, Qwen, DeepSeek, Gemma, Phi, Anthropic, OpenAI and Cohere models, without the prompt ever crossing a border.
Browse the model catalogue → Built for Australian AI workloads
Sovereign by design. Compatible by default.
Drop Amaze into your existing AI stack. Sovereign Australian inference, AUD pricing, and a sovereign Australian endpoint behind every model.
Zero retention
Prompts and outputs discarded by default. Nothing cached across tenants.
Burst GPU access
Scale into dedicated GPU pools when batch jobs and fine-tunes demand it.
Carrier-grade transit
Up to 40 Gbps onto the Amaze backbone. Predictable latency, free inter-region traffic.
Why Amaze for AI Sovereignty isn't a feature flag.
Sovereignty isn't a feature flag.
It's the architecture.
Amaze runs AI workloads exclusively on Australian-controlled infrastructure under Australian law. Built on enterprise Dell hardware, OpenStack, Fortinet security and Equinix / NEXTDC data centres in Sydney and Melbourne.
Sovereign by design
Inference never leaves Australia. No prompt caching, no training on your data, no foreign control plane.
100 % AU-hosted
Faster to value
Rapid deployment based on custom requirements. Australian onboarding, Australian support, AUD invoicing, no overseas tickets, no time-zone tax.
AU onboarding
Compliance, in the open
ISO 27001 certified controls in production. Controls mapped to Australian regulatory frameworks.
ISO 27001 certified
Security & compliance
Compliance, in the open.
The frameworks that govern how we run, store and handle data. Audit reports available under NDA.
ISO 27001
CertifiedInformation security management system covering Amaze production infrastructure in Sydney and Melbourne.
Tier 3 Data Centres
AustralianEquinix and NEXTDC Tier 3+ facilities in Sydney and Melbourne. Carrier-neutral, 99.9% facility uptime.
Privacy Act 1988
CompliantAustralian privacy principles enforced across data handling. Notifiable Data Breach scheme aligned.
Run AI workloads right here in Australia.
The latest open-weight models, hosted on Amaze GPU pools in Australia.
AI Model catalogue → Qwen Qwen3.6-35B-A3B
Qwen Qwen3-235B-A22B
Qwen Qwen3-Coder-480B
DeepSeek DeepSeek V4-Pro
DeepSeek DeepSeek R1-0528
Google Gemma 3 27B
Mistral Mistral Small 3.2
Kimi Kimi K2.5
Kimi Kimi K2 Thinking
Z.ai GLM-5.2
Microsoft Phi-4 Reasoning
All vendors + 23 more →
Model catalogue
Frontier and open-weight, side by side.
Llama, Mistral, Qwen, DeepSeek, Gemma and Phi for open-weight inference. Anthropic Claude, OpenAI GPT-class, and Cohere Command for frontier work. One sovereign contract, one Australian support line, one set of compliance controls.
Stable APIs, AUD pricing, no forced model upgrades. If a model is deprecated upstream, you get a migration path with months of notice, not a Friday-afternoon EOL email.
Australian infrastructure built with leading global technology
Multimodel Inferencing is coming soon.
Sovereign Australian inference across the latest open-weight models is in the final stages of build. Get in touch to be first in line for early access when we go live.