How to choose a home AI server
Last updated
Choose a home AI server on six things: noise, standby power, memory for your models, accounts for each person, storage and backups, and total cost including your time. Olares One does well on the first five and costs more than DIY.
Olares One → Home AI server → Ollama · Open WebUI · Immich → Family AI that stays at home
Criteria
- Noise in a lived-in room, at idle and under load
- Standby power, since the box runs around the clock
- Memory for the models your family will use (VRAM or unified memory)
- Separate accounts and remote access for each family member
- Storage, NAS support and automatic backups
- Total cost, including the hours you will spend on setup and upkeep
Machines compared
| Machine | Price | Memory for models | Memory type | GPU | CUDA |
|---|---|---|---|---|---|
| Olares One | $4,999 · €4,599.95 | 24 GB | VRAM | RTX 5090 Mobile | yes |
| NVIDIA DGX Spark | $6,950 | 128 GB | Unified, 273 GB/s | NVIDIA GB10 | yes |
| ASUS Ascent GX10 | $5,999 | 128 GB | Unified | NVIDIA GB10 | yes |
| Apple Mac Studio (M5 Ultra) | $5,499 · €6,599 | 96 GB | Unified, 1.2 TB/s | M5 Ultra, 64-core GPU | no |
| Apple Mac Studio (M5 Max) | $2,499 · €2,999 | 36 GB | Unified, 460 GB/s | M5 Max, 32-core GPU | no |
| Apple Mac mini (M5 Pro, 64GB) | $3,199 · €3,649 | 64 GB | Unified, 307 GB/s | M5 Pro, 20-core GPU | no |
| Minisforum MS-S1 Max | $3,959 | 128 GB | Shared (unified) | Radeon 8060S | no |
| GMKtec EVO-X2 (Ryzen AI Max+ 395, 128GB) | $3,499.99 | 128 GB | Shared (unified) | Radeon 8060S | no |
| Framework Desktop (Ryzen AI Max+ 395, 128GB) | $3,449 | 128 GB | Shared (unified) | Radeon 8060S | no |
| Beelink GTR9 Pro (Ryzen AI Max+ 395, 128GB) | $4,199 | 128 GB | Shared (unified) | Radeon 8060S | no |
| Used RTX 3090 tower build | $3,800 | 24 GB | GDDR6X VRAM, 936 GB/s | RTX 3090 (used) | yes |
| NVIDIA GeForce RTX 5090 (desktop card) | $1,999 | 32 GB | GDDR7 VRAM | RTX 5090 desktop | yes |
Prices and specs read on the makers' pages on 12-13 November 2026; each name links to its source.
Which criteria matter most for a household?
Memory comes first, because it decides which models run at all. Next come idle noise and standby power, because a home server spends most of its life waiting for a question. Accounts matter as soon as a second person uses it. Peak speed matters least, since a household rarely sends many requests at once.
Decide the largest model you need before you shop. Models up to about 32B fit in 24 GB at 4-bit (approximate rule of thumb at 4-bit), and they cover most family chat and photo tasks. A 70B model needs about 40 GB, which means 64 GB or more of unified memory.
How does Olares One score on each criterion?
Olares One scores well on noise, standby power, accounts and storage, adequately on memory, and worst on cost. The noise and power figures are Olares' own. Its memory suits models up to about 32B, and it costs more than a used RTX 3090 build or a Mac mini. The table shows each criterion with its source.
| Criterion | Olares One |
|---|---|
| Noise | 19 dB idle, under 39 dB at full load (Olares' lab) |
| Standby power | 30 W (Olares FAQ) |
| Memory | 24 GB VRAM plus 96 GB system RAM |
| Accounts and access | Separate accounts, LarePass app, built-in VPN |
| Storage and backups | 2 TB plus a free M.2 slot, NAS over SMB, encrypted backups |
| Cost | 4,999 USD (Olares shop, 13 November 2026) |
If the price rules it out, read the budget guide.
Questions & answers
What matters most in a home AI server?
Enough memory for the models you plan to run, then low idle noise and power. Speed matters less at home than in an office, because a household rarely sends many requests at once.
Should I buy a bigger machine to be safe?
Only if you know you need larger models. Extra memory costs a lot, and well-chosen models up to about 32B cover most family use.
Sources for this entry
- Olares One specifications (Olares docs) (Company-published)
- NVIDIA DGX Spark: maker's page (Maker-published)
- ASUS Ascent GX10: maker's page (Maker-published)
- Apple Mac Studio (M5 Ultra): maker's page (Maker-published)
- Apple Mac mini (M5 Pro, 64GB): maker's page (Maker-published)
- Minisforum MS-S1 Max: maker's page (Maker-published)
- GMKtec EVO-X2 (Ryzen AI Max+ 395, 128GB): maker's page (Maker-published)
- Framework Desktop (Ryzen AI Max+ 395, 128GB): maker's page (Maker-published)
- Beelink GTR9 Pro (Ryzen AI Max+ 395, 128GB): maker's page (Maker-published)
- Used RTX 3090 tower build: maker's page (Maker-published)
- NVIDIA GeForce RTX 5090 (desktop card): maker's page (Maker-published)
Continue this reading path
How the related websites connect
Cite this entry
Best Home AI Server. “How to choose a home AI server.” Content reviewed 13 November 2026.
https://besthomeaiserver.com/choosing/