4 bays · 2 x 2.5GbE
AINAS Score 23
No verified price

AI CapableAoostar
Why AI Capable: Ryzen 7 5825U (no NPU), no GPU expansion path, and Aoostar does not publish a memory ceiling (NASCompares says 64GB), so capable on manufacturer data.
Updated
Can the Aoostar R7 run local AI?
Partly. The Aoostar R7 runs local models on AMD Ryzen 7 5825U alone, and its nullGB ceiling allows only very small models of around 1B to 3B parameters (Estimated), but replies will be slow without graphics acceleration.
No graphics card can be fitted, which keeps every model on the CPU.
Local LLM support is confirmed by community reports.
The R7 suits a buyer who wants a compact two-bay box that runs Windows 11 Pro and containers, and who treats AI as a light extra. It has an AMD Ryzen 7 5825U with 8 cores, two SATA bays, two M.2 slots and two 2.5GbE ports. Docker and virtual machines are both supported, and a Windows 11 Pro key is included, so there is no vendor NAS software to learn. Memory is DDR4 3200 SO-DIMM in dual channel and is expandable, so you choose the capacity. It is sold barebone, with no memory or storage included. Ollama has a native Windows build (community confirmed), but inference is CPU-only. The AINAS Score is 22. It draws from a 90W external power supply.
The thing that rules out the R7 for some buyers is the lack of a GPU path combined with an unpublished memory ceiling. Aoostar does not state a maximum, and the R7 has no OCuLink, USB4 or PCIe slot, so graphics cannot be added later. This affects anyone who wants to run larger models locally. Look for a system with a stated high memory ceiling, a GPU route such as OCuLink or a PCIe slot, or an NPU. Estimated: none of the models listed on this page fit on the R7 as sold, at Q4 or Q8, because it ships with no memory. Buyers who only want a Windows mini server with Docker are not affected by this.
No verified price
We have not verified a listing for the Aoostar R7 at a retailer we track. The specifications below stand on their own; a price appears here as soon as a listing is confirmed.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon.co.uk or Amazon.com, as applicable, at the time of purchase will apply to the purchase of this product.
Each capability shows how it was confirmed: Verified by AINAS, Vendor confirmed, Community confirmed or Unverified.
| Capability | Supported | Evidence | Note |
|---|---|---|---|
| Local LLM | Limited | Community confirmed | No vendor AI software. Ships with Windows 11; Ollama has a native Windows build. CPU-only inference. |
| Ollama | Yes | Community confirmed | Ollama offers a native Windows installer; Linux NAS OS installs are user choice. |
| Open WebUI | Yes | Unverified | Runnable via Docker or Python on Windows/Linux; no platform specific source found. |
| RAG | Limited | Unverified | Only via user installed tools such as Open WebUI; no vendor feature. |
| AI search | Limited | Unverified | No vendor NAS OS or AI search; third party apps only. |
| GPU expansion | No | Vendor confirmed | No OCuLink, USB4 or PCIe slot. |
| Hardware acceleration | Limited | Unverified | Radeon iGPU not listed on R7 page; no NPU. |
| AI photos | Limited | Unverified | No vendor photo app; third party apps such as Immich via Docker not verified on this hardware. |
| AI video | Limited | Unverified | No vendor feature; third party only, not verified. |
| Surveillance AI | Limited | Unverified | No vendor surveillance app; third party software only, not verified. |
Calculated from the R7's memory and each model's published size at each quantization. Rows marked Estimated are calculations, not benchmarks. Check any model.
| Model | Best quantization | Runs? | Memory needed | Expected performance | Verification |
|---|---|---|---|---|---|
| Llama 3.2 3BLlama, 3.21B | None fits | No | 11GB at Q4 | Not measured | Estimated |
| Llama 3.1 8BLlama, 8.03B | None fits | No | 14GB at Q4 | Not measured | Estimated |
| Qwen3 8BQwen, 8.19B | None fits | No | 15GB at Q4 | Not measured | Estimated |
| Gemma 3 12BGemma, 11.8B | None fits | No | 18GB at Q4 | Not measured | Estimated |
| Qwen3 14BQwen, 14.8B | None fits | No | 20GB at Q4 | Not measured | Estimated |
| Phi-4 14BPhi, 14.7B | None fits | No | 19GB at Q4 | Not measured | Estimated |
| Mistral Small 3 24BMistral, 23.6B | None fits | No | 25GB at Q4 | Not measured | Estimated |
| Gemma 3 27BGemma, 27B | None fits | No | 29GB at Q4 | Not measured | Estimated |
| Qwen3 30B A3BQwen, 30.5B | None fits | No | 31GB at Q4 | Not measured | Estimated |
| Qwen3 32BQwen, 32.8B | None fits | No | 32GB at Q4 | Not measured | Estimated |
| DeepSeek-R1 32B (Qwen distill)DeepSeek, 32.8B | None fits | No | 32GB at Q4 | Not measured | Estimated |
| Llama 3.3 70BLlama, 70.6B | None fits | No | 60GB at Q4 | Not measured | Estimated |
| Llama 4 Scout 17Bx16ELlama, 109B | None fits | No | 89GB at Q4 | Not measured | Estimated |
| Qwen2.5 72BQwen, 72.7B | None fits | No | 65GB at Q4 | Not measured | Estimated |
| Mistral Large 2 123BMistral, 123B | None fits | No | 96GB at Q4 | Not measured | Estimated |
One verdict per workload, from the R7's use-case scores and capability rows. Unknown means we cannot confirm it, not that it probably works.
Poor fit for running a local LLM. Ollama is community confirmed on Windows, but the local LLM row is limited, with CPU-only inference and no vendor AI software. No listed model fits at Q4 or Q8 (Estimated).
Poor fit for private RAG on the R7. RAG is limited and Unverified, possible only through user installed tools such as Open WebUI, which is itself listed as runnable via Docker but Unverified, with no vendor feature.
We cannot confirm AI search on the R7. The capability is limited and Unverified, with no vendor NAS OS or AI search feature, so only third party apps are possible, and none is confirmed.
Unknown: we cannot confirm AI photos on the R7. There is no vendor photo app, and third party apps such as Immich via Docker are not verified on this hardware, so support is limited and Unverified.
Unknown: we cannot confirm surveillance use on the R7. There is no vendor surveillance app, and third party software is not verified, so the capability is limited and Unverified; Docker is supported, but no camera software is confirmed on it.
Poor fit for a home lab, despite Docker and virtual machine support. The R7 has no GPU path, no NPU and no stated memory ceiling, and its two 2.5GbE ports are the fastest networking it offers.
Poor fit for business use on the R7. It has no ECC memory, a one-year warranty and no vendor NAS OS, only a Windows 11 Pro key, so there is little to support a managed deployment.
Poor fit for creative AI. There is no GPU expansion (vendor confirmed, no OCuLink, USB4 or PCIe slot), hardware acceleration is limited and Unverified, and Aoostar does not list the Radeon iGPU on the R7 page, with no NPU either.
The R7 supports up to an unstated amount ofGB of DDR4 3200 SO-DIMM (dual channel).
There is no supported way to add a GPU to the R7: no PCIe slot and no OCuLink. Inference runs on the CPU, so keep to small models.
For a personal library, yes. A compact embedding model and a small chat model will run, but the R7 is better as the document store behind a RAG system whose model runs on stronger hardware.
Built-in networking: 2 x 2.5GbE. Adequate for a household, slower for bulk indexing.
| Configuration | RAM | SSD | Notes |
|---|---|---|---|
| Barebone | None | None | Free Windows 11 Pro key. |
The closest Aoostar model by bay count and AINAS Score. Only the specifications that differ are listed.
Its strongest AINAS profiles are Business at 25 and Home Lab at 24, which points to small offices that need shared storage with private AI tools and room to grow.
Look elsewhere if you are anyone who wants to run models larger than about 8B parameters, or buyers who expect to add a graphics card later.
Other AI Capable systems closest in overall score.
No. Estimated: Llama 3.2 3B needs 2GB at Q4 and 3.4GB at Q8, and the estimate says it does not fit at either, because the R7 comes barebone with no memory included. This is a calculation, not a benchmark. Ollama has a native Windows build (community confirmed), but local LLM support is limited to CPU-only inference. The same estimate gives no for every model listed on this page.
Yes. Memory is DDR4 3200 SO-DIMM in dual channel and is expandable, and the R7 is sold barebone with no memory included, so you fit your own. Aoostar does not publish a maximum capacity, so the ceiling is unconfirmed by the vendor and you should check before buying large modules. The R7 does not support ECC memory, so choose standard non-ECC modules.
No. Aoostar lists no OCuLink, USB4 or PCIe slot, and the GPU expansion row is vendor confirmed as no. The ports are USB 3.2 Gen 2 and USB 2.0 only. The R7 also has no NPU, so AI work runs on the CPU. Hardware acceleration is limited and Unverified, and Aoostar does not list the Radeon iGPU on the R7 page.
Yes. Docker and virtual machines are both supported. The R7 comes with a Windows 11 Pro key rather than a vendor NAS OS, and the Aoostar page lists Windows 10/11, OPNsense, OpenWrt and Ubuntu. Open WebUI is described as runnable via Docker or Python, but that is Unverified on this hardware. Ollama has a native Windows installer (community confirmed).
One, the barebone. It includes a free Windows 11 Pro key but no memory, no M.2 drive and no SATA disks, so the buyer supplies all three. The memory is DDR4 3200 SO-DIMM in dual channel without ECC, and the two M.2 slots take 2280 NVMe drives at PCIe Gen 3. Power comes from a 90W external supply, and the warranty is 1 year. Because no memory is fitted, every model fit on this page reads as no until you add modules (Estimated).
Yes, two. The R7 has two M.2 slots alongside two 2.5/3.5in SATA bays. The slot count is medium confidence, because Aoostar's wording could mean two slots or x2 lanes. A community review says the R7 has a second M.2 Gen 3 slot. Storage is not included, so you supply your own drives.
Yes. Ollama support is confirmed by community reports. See the capability table above for the source.
Yes. Open WebUI support is not yet verified. See the capability table above for the source.
Partly. AI photos support is not yet verified. See the capability table above for the source.
Partly. Surveillance AI support is not yet verified. See the capability table above for the source.
Partly. Hardware acceleration support is not yet verified. See the capability table above for the source.
It has 2 x 2.5GbE built in. The fastest port is 2 x 2.5GbE.
Every field records its source. Specifications were last checked on . Spot an error? Send a correction with a manufacturer link.