M6Temporarily out of stock
Mac mini M6 12CPU 12GPU 16GB 256GB
- Memory16GB
- Memory bandwidth153GB/s
- Estimated model sizeup to ~19B parameters (4-bit)
- Estimated ready dateTemporarily out of stock
4,990,000₫/month · incl. 10% VAT
Out of stockNamtech provides dedicated AI compute on Mac Studio and Mac mini for each business, with managed service. The machine stays at Namtech, each client gets a dedicated machine, and data is kept separate from other clients. You call the model over an API (optionally over VPN to the API), and Namtech handles setup, monitoring and maintenance. 12-month prices from 10,490,000₫/month, including 10% VAT.
The machine lives at Namtech. You only need a network connection to use it, with no power, cooling or software updates to worry about. The service is for running AI models only: you call the model over an API, you do not get remote control of the machine, and it is not used for anything else.
Choose a configuration on a 12-month term. Pay the first period in advance plus a deposit.
Namtech installs a suitable model (Qwen, Gemma, GLM…) using MLX, Ollama or LM Studio, based on your needs.
You receive an API address and access key for the model on your dedicated machine, reachable over VPN if needed. Only authorised users can call the model.
Namtech monitors the machine, applies updates and maintenance. If something goes wrong, Namtech handles it.
12-month term. Prices include initial setup, power, monitoring and maintenance.
Prices include 10% VAT. Model sizes are RAM-based estimates to help you choose, not benchmark results.
Each option has different costs and limits. The biggest difference is memory available to the model: a 24GB GPU and a large-RAM Mac run different model sizes.
| Option | Cost | Memory for the model | Estimated model size | Who operates it |
|---|---|---|---|---|
| Rent M5 Max 36GB | 10,490,000₫/month | ~27GB (unified memory) | ~43B | Namtech |
| Rent M5 Max 48GB | 12,490,000₫/month | ~36GB (unified memory) | ~57B | Namtech |
| Rent M5 Ultra 256GB | 31,990,000₫/month | ~192GB (unified memory) | ~307B | Namtech |
| Buy M5 Max 48GB | 98,610,000₫ (one-off, incl. VAT) | ~36GB (unified memory) | ~57B | Your team |
| Cloud GPU VNDATA RTX A5000 24GB | 8,000,000₫/month | 24GB VRAM | ~27B | Provider runs the server, you install the model |
| Cloud GPU FPT Cloud A30 24GB | 11,800,000₫/month | 24GB VRAM | ~27B | Provider runs the server, you install the model |
Each client is assigned a dedicated machine. One client’s data and models never sit on the same machine as another’s.
Under Vietnam’s Personal Data Protection Law No. 91/2025/QH15, you are the data controller and Namtech is the data processor, processing only under the contract and your instructions.
Only accounts you authorise can call the model. Namtech staff access the machine for maintenance under defined permissions, with records kept.
When the contract ends, Namtech deletes the data on the machine and sends you a deletion record before the machine is reused.
No system is perfectly secure. Specific security measures, each party’s responsibilities and incident handling are set out in the contract and data processing annex.
Service fees are paid before each period.
A deposit is paid at signing and settled at the end of the contract as per its terms.
Namtech issues a VAT invoice for each payment period. Payment is by bank transfer.
If payment is overdue, the service may be suspended (access is locked; your data is kept) until it is paid in full.
Service provider: N.A.M Technology Co., Ltd. (CÔNG TY TNHH CÔNG NGHỆ N.A.M, Namtech), tax code 0313959901. This is a service contract, not an equipment lease; the machine is owned by Namtech and is not delivered to you. The deposit amount and detailed terms are set out in the contract.
12-month rental starts from 10,490,000₫/month including 10% VAT, depending on the Mac mini or Mac Studio configuration.
No. The machine stays at Namtech. You call the model remotely via an API (optionally over VPN), and Namtech manages the machine.
No. The service is for running AI models only. Namtech does not provide remote control of the machine (screen or command line), and the machine is not used for rendering, software builds or other workloads.
No. Each client is assigned a dedicated machine. When the contract ends, Namtech deletes the data and issues a deletion record.
Open models such as Qwen, Gemma and GLM, run with MLX, Ollama or LM Studio. The largest model size depends on the RAM of the rented configuration.
A 24GB VRAM GPU fits a model of about 27B parameters at 4-bit. A large-RAM Mac can run larger models, and Namtech installs the model and operates the machine for you.
Paid in advance monthly or quarterly, with a deposit, by bank transfer with a VAT invoice. If payment is overdue, the service may be suspended.
Yes. Prices listed on this page include 10% VAT.
Tell us what you plan to use AI for and how many people will use it. Namtech will recommend a configuration and term that fit.