32GB max RAM · 2 bays · 2 x 2.5GbE
AINAS Score 29
No verified price

AI CapableQNAP
Why AI Capable: Intel Celeron N4505 (2C/2T) with no NPU, Intel UHD Graphics only, a PCIe Gen3 x2 slot with no GPGPU entries on QNAP's compatibility list, and a 16GB memory ceiling: Docker and small models only.
Updated
Can the QNAP TS-262 run local AI?
Only lightly. Expect 7B to 8B class models (Estimated) on the QNAP TS-262: Intel Celeron N4505 (2-core/2-thread, burst up to 2.9 GHz) does the work with up to 16GB of memory, which is fine for experiments rather than a daily assistant.
Without any GPU option, the processor carries the whole inference load.
Local LLM support is documented by the manufacturer.
The TS-262 suits a buyer who wants a small two-bay QNAP NAS for files, containers and light AI helpers, and who accepts that local language models are a side use. It has an Intel Celeron N4505 with 2 cores, 4GB of DDR4 SODIMM memory that expands to 16GB, two M.2 slots and one 2.5GbE port. It runs QTS 6.1.0 with Docker and virtual machines. QNAP confirms Ollama and Open WebUI tutorials, QuMagie photo management and QVR surveillance. Small models fit once memory is upgraded: Qwen3 8B fits at Q4 (Estimated). Intel UHD Graphics handles hardware transcoding, and an optional Coral Edge TPU of 4 TOPS can be added through M.2 or USB. The AINAS Score is 29.
The 16GB memory ceiling and the lack of any GPU path rule the TS-262 out for anyone who wants mid-size or large models. The PCIe Gen3 x2 slot has no GPUs on QNAP's compatibility list, and there is no NPU. Qwen3 32B needs 20GB at Q4, so it does not fit (Estimated). Buyers in that position should look for a system with a route to a graphics card, such as a full-width PCIe slot, OCuLink or Thunderbolt, and a memory ceiling well above 16GB. Included memory is a second concern. The TS-262-4G comes with 4GB, so every Estimated fit above assumes both SODIMM slots are filled to reach 16GB, and that memory has no ECC. The 2-core Celeron N4505 then does all the work on its own.
No verified price
We have not verified a listing for the QNAP TS-262 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 | Vendor confirmed | QNAP Ollama guide covers x86 QTS models (8GB for 3B, 16GB for 7B). 2-core Celeron and 16GB max: tiny models only, needs RAM upgrade from 4GB. |
| Ollama | Yes | Vendor confirmed | Container Station Ollama tutorial for x86 QTS/QuTS hero NAS. |
| Open WebUI | Yes | Vendor confirmed | QNAP Open WebUI tutorial, QTS 5.2 or later; model runs QTS 6.1.0. |
| RAG | Limited | Unverified | Qsirch RAG Search (beta) exists but requirements not stated; not confirmed for this model. |
| AI search | Limited | Unverified | Qsirch AI Semantic Search needs 64-bit x86 and 8GB RAM (ships with 4GB); AI Semantic Index requires x74 series or above, which this x62 model is not. |
| GPU expansion | No | Vendor confirmed | PCIe Gen3 x2 slot; no GPUs on the official compatibility list. |
| Hardware acceleration | Yes | Vendor confirmed | UHD Graphics hardware transcoding, OpenVINO for image recognition, optional Coral Edge TPU (4 TOPS) via M.2 or USB. |
| AI photos | Yes | Vendor confirmed | QuMagie AI-assisted photo management. |
| AI video | Yes | Vendor confirmed | QVR Face and QVR Human listed. |
| Surveillance AI | Yes | Vendor confirmed | QVR Surveillance 2 free + 6 on registration; QVR Pro 8 free. |
Calculated from the TS-262'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 TS-262's use-case scores and capability rows. Unknown means we cannot confirm it, not that it probably works.
The TS-262 is a poor choice for running local LLMs. QNAP's Ollama guide is vendor confirmed, but local LLM support is rated limited: the notes say tiny models only, with a 2-core Celeron and a memory upgrade from 4GB needed.
Private RAG is a poor fit here. Qsirch RAG Search is in beta, its requirements are not stated, and it is unconfirmed for this model, while Open WebUI and Ollama are vendor confirmed but sit on limited hardware.
We cannot confirm AI search on the TS-262. Qsirch AI Semantic Search needs 8GB of RAM and the model ships with 4GB, and the AI Semantic Index needs a higher series than this model, so support is unverified.
AI photos is rated poor on this system, despite a confirmed app. QuMagie AI-assisted photo management is vendor confirmed, and OpenVINO image recognition is listed with the Intel UHD Graphics, but the CPU has only 2 cores.
Surveillance is poor on the TS-262, though the software is present. QVR Surveillance includes 2 free plus 6 on registration, QVR Pro 8 is free, and QVR Face and QVR Human are listed, but the verdict reflects the 2-core CPU and 4GB of included RAM.
Home lab use is poor, although the platform handles the basics. It runs Docker and Virtualization Station on QTS 6.1.0, which needs 4GB minimum for virtual machines, but a 2-core CPU and a 16GB memory ceiling cap how many workloads you can run.
Business use is poor on the TS-262. It has two drive bays, no ECC memory, one 2.5GbE port (faster networking needs a PCIe card), a 2-year warranty, and a 2-core Celeron N4505.
Creative AI is poor on the TS-262. GPU expansion is vendor confirmed as unsupported, with no GPUs on QNAP's compatibility list for the PCIe Gen3 x2 slot, and local LLM support is only limited, with only UHD Graphics acceleration available.
Plan on upgrading. 4GB is fitted as standard, while the TS-262 accepts 16GB of DDR4 SODIMM, which is the difference between only very small models of around 1B to 3B parameters and 7B to 8B class models (Estimated).
Our estimate for the biggest model that fits in 16GB is about 13B parameters at 4-bit, after setting aside 8GB for the system. Long context windows reduce that figure (Estimated).
A graphics card is not an option on the TS-262, which has neither a PCIe slot nor OCuLink. Inference runs on the CPU, with Intel UHD Graphics handling media tasks, so keep to small models.
As the storage half of a RAG setup it works well; as the model host it is limited. Small embedding models run fine on the TS-262, while the answering model needs more compute than it has.
For network access it has 2.5GbE. Enough for light use; large imports will take a while.
| Configuration | RAM | SSD | Notes |
|---|---|---|---|
| TS-262-4G | 4GB | None |
The closest QNAP model by bay count and AINAS Score. Only the specifications that differ are listed.
Best fit: small offices that need shared storage with private AI tools and room to grow. That is where the TS-262 scores highest (Business 32/100), followed by Home Lab (31/100).
Less suited to anyone who wants to run models larger than about 8B parameters and buyers who expect to add a graphics card later.
Other AI Capable systems closest in overall score.
Yes, once the memory is upgraded. Qwen3 8B needs 5.2GB at Q4 and fits at Q4 (Estimated). At Q8 it needs 8.9GB and is marginal (Estimated). The TS-262 ships with 4GB and takes up to 16GB. QNAP's own notes say tiny models only, and local LLM support is rated limited. These are calculations, not benchmarks.
No. QNAP's compatibility list for the PCIe Gen3 x2 slot has no GPUs, and GPU expansion is vendor confirmed as unsupported. The slot takes other cards, such as 5GbE or 10GbE network cards, within a card limit of 190 x 68.9 x 18.76mm. The system has no OCuLink, USB4 or Thunderbolt ports either. The only AI accelerator option is an optional Coral Edge TPU of 4 TOPS through M.2 or USB.
16GB. It has two DDR4 SODIMM slots (2 x 8GB) and ships with 4GB, so the memory is upgradeable. It does not support ECC memory. Estimated model fits depend on this: Qwen3 8B fits at Q4 (Estimated), Gemma 3 12B is marginal at Q4 (Estimated), and Mistral Small 3 24B does not fit at Q4 (Estimated). Marginal means it loads with little headroom.
Only just, and only after a memory upgrade. Gemma 3 12B needs 8.2GB at Q4 and is marginal on the TS-262 at 16GB (Estimated), which means it loads with little headroom. At Q8 it needs 13GB and does not fit (Estimated). The 4GB the TS-262-4G ships with is not enough for either, so both SODIMM slots need filling first. With no GPU path and no NPU, the 2-core Celeron N4505 does all the work. These are calculations, not benchmarks.
Yes. QTS 6.1.0 includes Container Station, which runs LXC/LXD and Docker, and Virtualization Station is listed for virtual machines with a 4GB RAM minimum. QNAP's Ollama and Open WebUI tutorials are vendor confirmed, and Open WebUI needs QTS 5.2 or later. The 2-core Celeron and 4GB of included memory are the limits on how much runs at once.
Two. The TS-262 has 2 x M.2 2280 PCIe Gen3 x1 slots, alongside its 2 drive bays. Drive bays and M.2 slots are separate. Hardware acceleration is vendor confirmed, and QNAP lists an optional Coral Edge TPU of 4 TOPS that connects through M.2 or USB. That makes M.2 an option for an accelerator as well as storage. There is no NPU built into the TS-262.
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.
Yes. Hardware acceleration support is documented by the manufacturer. See the capability table above for the source.
It has 2.5GbE built in, and the 1x PCIe Gen3 x2 (card max 190 x 68.9 x 18.76 mm) can take a faster network card. The fastest port is 2.5GbE.
Every field records its source. Specifications were last checked on . Spot an error? Send a correction with a manufacturer link.