Skip to content
Matilda TrainSign up

GPU pricing in Australia. One rate per GPU-hour, everything included.

AMD Instinct MI355X training is A$9.35 per GPU-hour including GST on demand, billed by the second, or as little as A$4.95 including GST on a reserved node. Prices are in Australian dollars including GST, with data transfer and a Maincode data scientist included.

Sign up

Rates by booking term

Reserve whole 8-GPU nodes for less, with nothing upfront. Preemptible capacity suits jobs that checkpoint.

AMD Instinct MI355X. Australian dollars including GST, with ex-GST hourly rates shown alongside. Node prices use 8 GPUs × 730 hours. Reserved terms are paid in advance; on-demand monthly cost is illustrative.
BookingPer GPU-hourPer node, monthlyReserved term totalGood for
On demandA$9.35 incl GST
A$8.50 ex GST		A$54,604.00–Experiments and bursts, billed per second
1 monthA$9.075 incl GST
A$8.25 ex GST		A$52,998.00A$52,998.00A single training run
3 monthsA$8.58 incl GST
A$7.80 ex GST		A$50,107.20A$150,321.60A project
6 monthsA$8.25 incl GST
A$7.50 ex GST		A$48,180.00A$289,080.00A research program
12 monthsA$7.48 incl GST
A$6.80 ex GST		A$43,683.20A$524,198.40Ongoing training
24 monthsA$6.545 incl GST
A$5.95 ex GST		A$38,222.80A$917,347.20A model roadmap
36 monthsA$4.95 incl GST
A$4.50 ex GST		A$28,908.00A$1,040,688.00The lowest rate
PreemptibleA$4.675 incl GST
A$4.25 ex GST		––Jobs that checkpoint and can be requeued

Every booking includes

  • The MCP endpoint for your coding agents
  • Managed Slurm, SSH and a REST API
  • Data transfer in and out
  • A persistent workspace for code, data and checkpoints
  • PyTorch, Transformers, PEFT, Accelerate and Datasets, ready to go
  • A Maincode data scientist to port and tune your first runs

CPU-only allocations: A$0.066 per vCPU-hour including GST (A$0.06 ex GST), with up to 4 GiB RAM per vCPU included. Billed per second of allocated time. GPU allocations already include their CPU and memory.

NVIDIA B300: pricing on request. Tell us the run you have in mind and we’ll quote it.

What common runs cost

Worked from the rates above. Your own runs depend on the model, the data and how long training takes.

  • One 70B fine-tune

    A$12.69including GST

    MLPerf Llama 2 70B LoRA: 8 GPUs for 10.18 minutes, on demand

  • A day on one node

    A$1,795.20including GST

    8 GPUs × 24 hours × A$9.35, on demand

  • A node for a month

    A$43,683.20including GST

    8 GPUs × 730 hours × A$7.48, on a 12-month term

  • Two hours of CPU data prep

    A$2.11including GST

    16 vCPUs with 64 GiB RAM included × 2 hours × A$0.066, no GPU

One 70B fine-tune, priced in Australia

The GPU-hour price is half the story.
The other half is how long the run takes.

Cost of one run, Australian dollarsMLPerf Llama 2 70B fine-tune · 8 GPUs · on-demand prices excluding GST

Matilda TrainMI355X · 10.18 min

A$11.54

AWS SydneyH100 · 27.54 min

A$47.06

Google Cloud SydneyH100 · 27.54 min

A$64.43

Azure Australia EastH100 · 27.54 min

A$93.23

Vendor-submitted MLPerf Training results, times each provider’s on-demand GPU-hour price in its Australian region. Your workload will differ.

2.7× faster

than the best published 8× H100 result: 10.18 against 27.54 minutes

4× cheaper

per run than AWS Sydney on demand: A$11.54 against A$47.06, excluding GST

Pricing questions

More detail, examples and troubleshooting are in the researcher documentation.

How is billing worked out?
Per GPU-hour. On-demand use is billed per second at A$9.35 including GST (A$8.50 excluding GST), reserved nodes cost less, and preemptible capacity is A$4.675 including GST (A$4.25 excluding GST) for jobs that checkpoint. CPU-only allocations cost A$0.066 per vCPU-hour including GST (A$0.06 excluding GST), with up to 4 GiB RAM per vCPU included, billed per allocated second; GPU allocations include their CPU and memory. Prices are in Australian dollars. Transfer and support are included, subject to fair-use limits.
How do reserved nodes work?
Reserve whole 8-GPU MI355X nodes for 1 to 36 months, with the reserved term paid in advance. The rate falls with the term, from A$9.075 per GPU-hour including GST for one month to A$4.95 including GST for 36 months. Full term prices are shown in the table; contact Maincode to arrange a reserved booking. Monthly node prices are 8 GPUs × 730 hours.
What is preemptible capacity?
Capacity at A$4.675 per GPU-hour including GST (A$4.25 excluding GST) for jobs that checkpoint and can be requeued. Use it when a run can pick up again from its last checkpoint.
Do prices include GST?
Public prices are in Australian dollars and include Australian GST. Ex-GST hourly rates are shown separately. Your final purchase tax depends on your billing details. They include data transfer in and out, a persistent workspace and a Maincode data scientist for your first runs.
Do I own the model?
Yes. Checkpoints, adapters and merged weights are written to your workspace as standard files. Download them whenever you like and serve them anywhere. There’s no fee to keep them. The workspace guide covers downloads.
How much does NVIDIA B300 cost?
B300 pricing is on request. Email us about the run you have in mind and we’ll quote it.

Bring your training.
Own the model.

Create your account, add compute credits and connect your coding agent. You pay by the GPU-hour, and every checkpoint is yours to download.

Sign up Read the getting started guide

Amazon Bedrock, Google Cloud, Gemini Enterprise Agent Platform, Vertex AI, Gemini, Microsoft Foundry, Azure and other product names are trademarks of their owners, used here to identify and compare.