Hardware › Mac Studio M5 Ultra (80-core GPU) 256GB
Mac Studio M5 Ultra (80-core GPU) 256GB for local AI: price in Australia & what it can run
A$11,449cheapest found · Apple AU
256 GBunified (192.0 GB usable)
1,200 GB/smemory bandwidth
179 tok/smeasured, Llama 2 7B Q4_0
View at Apple AU → Price checked 20 Oct 2026, 7:44 am AEST.
Apple specs list 256GB or 512GB for the 36-core/80-core M5 Ultra; price shown is Apple's 'From' price for that chip.
Benchmark
179.1 tok/s on Llama 2 7B Q4_0, tg128 (llama.cpp llama-bench), Metal backend. Source: ggml-org/llama.cpp/discussions/4167 scoreboard (checked 2026-10-19).
Specs: https://www.apple.com/au/mac-studio/specs/
AU prices found
| Retailer | Listing | Price | Stock | Checked |
|---|---|---|---|---|
Apple AU | A$11,449 | in stock | 20 Oct 2026, 7:44 am AEST |
Community measured
No reader results yet. Be the first to submit one →
Which models fit? (8k context)
| Model | Quant | Needs / usable | Fit | Est. tok/s |
|---|---|---|---|---|
Q4_K_M | 464.4 / 192.0 GB | no | – | |
Q4_K_M | 499.4 / 192.0 GB | no | – | |
Q4_K_M | 20.0 / 192.0 GB | fits | 373 estimate | |
Q4_K_M | 348.5 / 192.0 GB | no | – | |
Q4_K_M | 15.7 / 192.0 GB | fits | 560 estimate | |
Q4_K_M | 45.9 / 192.0 GB | fits | 124 estimate | |
Q4_K_M | 22.5 / 192.0 GB | fits | 373 estimate | |
Q4_K_M | 305.3 / 192.0 GB | no | – | |
MiniCPM V 4.7 1B new | Q4_K_M | 2.0 / 192.0 GB | fits | 723 estimate |
Q4_K_M | 464.4 / 192.0 GB | no | – | |
Q4_K_M | 195.9 / 192.0 GB | no | – | |
GLM 5.3 (MoE) new | Q4_K_M | 458.4 / 192.0 GB | no | – |
Q4_K_M | 110.3 / 192.0 GB | fits | 187 estimate | |
Q4_K_M | 18.4 / 192.0 GB | fits | 40 estimate | |
Q4_K_M | 20.2 / 192.0 GB | fits | 373 estimate | |
Q4_K_M | 9.1 / 192.0 GB | fits | 94 estimate | |
Q4_K_M | 23.0 / 192.0 GB | fits | 373 estimate | |
Q4_K_M | 81.4 / 192.0 GB | fits | 9 estimate | |
Q4_K_M | 22.1 / 192.0 GB | fits | 36 estimate | |
Q4_K_M | 17.2 / 192.0 GB | fits | 295 estimate | |
Q4_K_M | 6.0 / 192.0 GB | fits | 140 estimate | |
Q4_K_M | 7.1 / 192.0 GB | fits | 116 estimate | |
Q4_K_M | 4.1 / 192.0 GB | fits | 240 estimate | |
Q4_K_M | 77.0 / 192.0 GB | fits | 112 estimate | |
Q4_K_M | 20.4 / 192.0 GB | fits | 373 estimate | |
gpt-oss 20B (MoE) older | MXFP4 | 12.4 / 192.0 GB | fits | 355 estimate |
gpt-oss 120B (MoE) older | MXFP4 | 63.5 / 192.0 GB | fits | 251 estimate |
Qwen 3 4B older | Q4_K_M | 4.6 / 192.0 GB | fits | 279 estimate |
Qwen 3 8B older | Q4_K_M | 7.2 / 192.0 GB | fits | 137 estimate |
Qwen 3 32B older | Q4_K_M | 23.0 / 192.0 GB | fits | 34 estimate |
Llama 3.1 8B Instruct older | Q4_K_M | 6.9 / 192.0 GB | fits | 139 estimate |
Mistral 7B Instruct v0.3 older | Q4_K_M | 6.5 / 192.0 GB | fits | 154 estimate |