32GB max RAM · 2 bays · 2 x 2.5GbE
AINAS Score 33
Price awaiting re-verification
Last checked

AI CapableASUSTOR
Why AI Capable: Intel Celeron N5105 (4C/4T) with Intel UHD Graphics, no NPU, 16GB memory ceiling and an expansion slot that only takes ASUSTOR's 10GbE card; small CPU-only models via ASUSTOR's Ollama app at best.
Updated
Can the ASUSTOR Lockerstor 4 Gen2 (AS6704T) run local AI?
The ASUSTOR Lockerstor 4 Gen2 (AS6704T) can run small local models, with limits. Its Intel Celeron N5105 and 16GB memory ceiling suit 7B to 8B class models (Estimated), so treat it as a storage-first NAS that can host light AI tools.
There is no way to add a discrete GPU, so all inference runs on the CPU and basic integrated graphics.
Local LLM support is documented by the manufacturer.
The Lockerstor 4 Gen2 is a four-bay desktop NAS for buyers who want storage first and small local AI models as an extra. It has four M.2 slots, two 2.5GbE ports, an Intel Celeron N5105 and 4GB of memory, expandable to 16GB. ASUSTOR's Ollama and Open WebUI app is vendor confirmed and installs through Docker, and the system also runs virtual machines. Photo Gallery with facial recognition and Surveillance Center are both vendor confirmed too. Model sizing is a calculation: Qwen3 8B fits at Q4 (Estimated), while 8B models at Q8 are marginal (Estimated). Inference is CPU-only. It carries an AINAS Score of 32 and a three-year warranty. Choose it as a capable file server with a light AI side.
The memory ceiling rules it out for anyone who wants to run larger models. The Lockerstor 4 Gen2 tops out at 16GB, and nothing in the system offers a GPU path: the only PCIe slot takes an optional 10GbE card. Mistral Small 3 24B does not fit at Q4 (Estimated), and neither do the 27B, 30B and 32B models listed on this page (Estimated). Buyers who plan to run models in that range should look for a system with memory well above 16GB, a GPU expansion path such as a PCIe slot that takes a graphics card, and ideally an NPU. Anyone who only needs files, photos and cameras can ignore this limit.
Amazon UK
4GB
£659.02
Verified
In stock
Buy on AmazonProduct 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 | Vendor confirmed | ASUSTOR App Central lists Ollama + Open WebUI (v0.11.4.r01, ADM 5.0+) for the AS67 series. CPU-only inference; 4GB base RAM and 16GB max limit model size. The app does not install a model by default to avoid failures on some NAS models. |
| Ollama | Yes | Vendor confirmed | Ollama + Open WebUI app lists AS67 as a supported series; installs via Docker. |
| Open WebUI | Yes | Vendor confirmed | Bundled with the Ollama + Open WebUI App Central package. |
| RAG | Limited | Vendor confirmed | ASUSTOR markets NAS as RAG knowledge base storage with vector DBs (Chroma, pgvector, Weaviate) via Docker; Open WebUI includes RAG. ASUSTOR recommends external compute for inference. |
| AI search | No | Unverified | No AI or semantic file search feature found in ADM 5.1 material. |
| GPU expansion | No | Vendor confirmed | Only expansion slot is for ASUSTOR's optional 10GbE card; no GPU path documented. ASUSTOR states its NAS models do not feature a discrete GPU. |
| Hardware acceleration | Limited | Vendor confirmed | Intel UHD Graphics iGPU (Intel ARK) used for 4K H.265 hardware decoding/transcoding per ASUSTOR; no NPU. GPU use for AI inference not documented. |
| AI photos | Yes | Vendor confirmed | Photo Gallery 4.2 (ADM 5.1.1+) lists AS67 series; ADM 5.1 adds facial recognition grouping powered by locally run AI. |
| AI video | No | Unverified | No AI video feature found in ADM documentation. |
| Surveillance AI | Yes | Vendor confirmed | Surveillance Center 2.8.4 (app 1663) and Surveillance Center 3 (app 1661) list AS67; up to 44 IP cameras per ASUSTOR spec page licence table. |
Calculated from the Lockerstor 4 Gen2 (AS6704T)'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 | Q8 | Yes | 13GB | Not measured | Estimated |
| Llama 3.1 8BLlama, 8.03B | Q5 | Yes | 15GB | Not measured | Estimated |
| Qwen3 8BQwen, 8.19B | Q4 | Yes | 15GB | Not measured | Estimated |
| Gemma 3 12BGemma, 11.8B | Q4 | Marginal | 15GB | Not measured | Estimated |
| Qwen3 14BQwen, 14.8B | Q4 | Marginal | 16GB | Not measured | Estimated |
| Phi-4 14BPhi, 14.7B | Q4 | Marginal | 16GB | 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 Lockerstor 4 Gen2 (AS6704T)'s use-case scores and capability rows. Unknown means we cannot confirm it, not that it probably works.
The Lockerstor 4 Gen2 is a poor choice for local LLM work. Local LLM support is limited to CPU-only inference through ASUSTOR's Ollama app (vendor confirmed), and 4GB base memory with a 16GB maximum caps model size.
Private RAG is a poor fit here. ASUSTOR lists vector databases via Docker and RAG in Open WebUI as limited support, and it recommends external compute for inference, since this system has no GPU path.
AI search is a poor fit on this system. No AI or semantic file search feature was found in ADM 5.1 material, so the capability is marked unverified and unconfirmed.
AI photos is rated poor overall for this system. Photo Gallery 4.2 lists the AS67 series, and ADM 5.1 adds facial recognition grouping powered by locally run AI (vendor confirmed).
Surveillance is rated poor overall for this system. Surveillance Center lists the AS67 series, with up to 44 IP cameras per ASUSTOR's licence table (vendor confirmed), but the Celeron N5105 has no NPU.
Home lab use is rated poor. Docker and virtual machines are both supported, but memory stops at 16GB, there is no ECC and the expansion slot only takes a 10GbE card.
Business use is rated poor. The system has four bays, two 2.5GbE ports and a three-year warranty, but no ECC memory and a 16GB ceiling limit heavier workloads.
Creative AI is a poor fit. GPU expansion is not supported (vendor confirmed), and the Intel UHD Graphics iGPU is documented for video transcoding only, not for AI inference.
The Lockerstor 4 Gen2 (AS6704T) ships with 4GB of DDR4-2933 SO-DIMM and supports up to 16GB. Upgrading to the maximum lifts the model ceiling from only very small models of around 1B to 3B parameters to 7B to 8B class models (both Estimated).
As a rule of thumb, a 4-bit model needs about 0.6GB per billion parameters plus room for context. After leaving around 8GB for the NAS operating system, 16GB leaves space for a model of roughly 13B parameters at most (Estimated).
No. The Lockerstor 4 Gen2 (AS6704T) has no PCIe slot or OCuLink port for a graphics card. Inference runs on the CPU, with Intel UHD Graphics handling media tasks, so keep to small models.
With limits. The Lockerstor 4 Gen2 (AS6704T) can host a small RAG stack such as AnythingLLM or Open WebUI with a compact embedding model, but memory and compute restrict how large the answering model can be. It suits a personal document library more than a team knowledge base.
Networking is 2 x 2.5GbE. Fine for one or two users; indexing a large archive over the network will be slower.
| Configuration | RAM | SSD | Notes |
|---|---|---|---|
| 4GB | 4GB | None | Diskless; 1x 4GB DDR4 pre-installed, 2 SO-DIMM slots. |
The closest ASUSTOR model by bay count and AINAS Score. Only the specifications that differ are listed.
The Lockerstor 4 Gen2 (AS6704T) scores best for Home Lab (38/100) and Business (37/100). It suits home lab owners who want one box for storage, containers, media and a few AI services.
It is a weaker fit for anyone who wants to run models larger than about 8B parameters and for buyers who expect to add a graphics card later.
Other AI Capable systems closest in price.
Yes, at Q4. Qwen3 8B needs 5.2GB at Q4, which fits within the 16GB maximum (Estimated). At Q8 it needs 8.9GB and is marginal (Estimated). This is a calculation, not a benchmark. Inference is CPU-only through ASUSTOR's Ollama app, and the app does not install a model by default. Gemma 3 12B at Q4 is marginal (Estimated).
No. The only expansion slot is a 1x slot for ASUSTOR's optional 10GbE card, and no GPU path is documented. ASUSTOR states its NAS models do not feature a discrete GPU. There is no OCuLink, Thunderbolt or USB4 port either. The Intel UHD Graphics iGPU handles 4K H.265 hardware decoding and transcoding, but its use for AI inference is not documented.
Yes. The system uses DDR4-2933 SO-DIMM memory and has two slots, with one 4GB module installed. The maximum is 16GB, as 2x 8GB. The memory is not ECC. ASUSTOR's compatibility list covers modules from 4GB to 16GB for this series, so check that list before buying a kit.
Yes. Docker Engine for this series needs ADM 5.1.0 or later, and ASUSTOR's Ollama and Open WebUI app installs through it, vendor confirmed. Virtual machines are supported too. For RAG, ASUSTOR points to vector tools such as Chroma, pgvector and Weaviate running in Docker, with external compute recommended for inference. With 4GB installed and a 16GB ceiling, containers and virtual machines share little memory, so a memory upgrade is the first step for anything beyond file services.
Yes, four. The Lockerstor 4 Gen2 has four M.2 NVMe slots alongside its four drive bays. It runs ADM 5, supports Docker and virtual machines, and has two 2.5GbE ports. The optional 10GbE card uses the PCIe slot, and ASUSTOR's press release says that card replaces the M.2 card.
Yes. Ollama support is documented by the manufacturer. See the capability table above for the source.
Yes. Open WebUI support is documented by the manufacturer. See the capability table above for the source.
Yes. AI photos support is documented by the manufacturer. See the capability table above for the source.
Yes. Surveillance AI support is documented by the manufacturer. See the capability table above for the source.
Partly. Hardware acceleration support is documented by the manufacturer. See the capability table above for the source.
It has 2 x 2.5GbE built in, and the 1x slot for optional 10GbE expansion card can take a faster network card. 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.