512GB max RAM · 24 bays · 2 x 10GbE
AINAS Score 61
No verified price

AI ReadyUGREEN
Why AI Ready: Intel Core 5 120U (10C/12T, no NPU) with Intel Iris Xe Graphics (80EU), internal PCIe Gen4 x4 slot and 2x Thunderbolt 4 for a GPU, and up to 96GB DDR5.
Updated
Can the UGREEN DXP6800 Ultra run local AI?
Partly. The UGREEN DXP6800 Ultra runs local models on Intel Core 5 120U alone, and its 96GB ceiling allows 70B class models at 4-bit with room left for longer context (Estimated), but replies will be slow without graphics acceleration.
A graphics card can be added through its PCIe x4 slot, which is the biggest single speed upgrade for local models.
Local LLM support is documented by the manufacturer.
The DXP6800 Ultra suits buyers who want a six-bay NAS that can also run local AI in Docker. It takes up to 96GB of DDR5 SO-DIMM 5200 memory in two slots, so the model list is long: Mistral Large 2 123B fits at Q4 (Estimated). Ollama and Open WebUI are both vendor confirmed on UGOS Pro. It has two 10GbE ports, two M.2 slots, an Intel Core 5 120U with 10 cores, and a 3-year warranty. The AINAS Score is 60.
The main catch is that inference on the DXP6800 Ultra is CPU-only. The local LLM row is marked limited, and there is no NPU. It ships with 8GB of memory, so you must buy more to use the larger models. Buyers who want fast chat with big models should look for a dedicated GPU path. The PCIe Gen4 x4 slot and the two Thunderbolt 4 ports provide one, and UGREEN documents dedicated GPU use in Docker. The GPU is a separate purchase, and you would need to check card size and power in the system before buying.
Amazon UK
8GB
£1,015.99£1,269.99
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 | CPU-only inference via Ollama in Docker per UGREEN KB; no NPU. |
| Ollama | Yes | Vendor confirmed | UGREEN KB Docker Compose guide for Ollama/Open WebUI on UGOS Pro. |
| Open WebUI | Yes | Vendor confirmed | Official UGOS Pro Open WebUI guide. |
| RAG | Limited | Vendor confirmed | RAG via Open WebUI document chat only. |
| AI search | Limited | Vendor confirmed | Photos description and OCR search via Model Manager; document semantic search applicability per model not stated. |
| GPU expansion | Yes | Vendor confirmed | Internal PCIe Gen4 x4 slot or Thunderbolt 4 eGPU; UGREEN KB documents dedicated GPU use in Docker. |
| Hardware acceleration | Limited | Vendor confirmed | Quick Sync on Iris Xe; CN page cites 8K transcoding; no NPU. |
| AI photos | Yes | Vendor confirmed | AI photo album: faces, places, objects, duplicates. |
| AI video | Limited | Vendor confirmed | Surveillance Center local AI detection (people, vehicle, package, pet); features vary by model. |
| Surveillance AI | Yes | Vendor confirmed | Surveillance Center supports DXP series. |
Calculated from the DXP6800 Ultra'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 | Q8 | Yes | 19GB | Not measured | Estimated |
| Qwen3 8BQwen, 8.19B | Q8 | Yes | 19GB | Not measured | Estimated |
| Gemma 3 12BGemma, 11.8B | Q8 | Yes | 24GB | Not measured | Estimated |
| Qwen3 14BQwen, 14.8B | Q8 | Yes | 28GB | Not measured | Estimated |
| Phi-4 14BPhi, 14.7B | Q8 | Yes | 28GB | Not measured | Estimated |
| Mistral Small 3 24BMistral, 23.6B | Q8 | Yes | 38GB | Not measured | Estimated |
| Gemma 3 27BGemma, 27B | Q8 | Yes | 44GB | Not measured | Estimated |
| Qwen3 30B A3BQwen, 30.5B | Q8 | Yes | 47GB | Not measured | Estimated |
| Qwen3 32BQwen, 32.8B | Q8 | Yes | 50GB | Not measured | Estimated |
| DeepSeek-R1 32B (Qwen distill)DeepSeek, 32.8B | Q8 | Yes | 50GB | Not measured | Estimated |
| Llama 3.3 70BLlama, 70.6B | Q5 | Yes | 68GB | Not measured | Estimated |
| Llama 4 Scout 17Bx16ELlama, 109B | Q4 | Yes | 89GB | Not measured | Estimated |
| Qwen2.5 72BQwen, 72.7B | Q5 | Yes | 73GB | Not measured | Estimated |
| Mistral Large 2 123BMistral, 123B | Q4 | Yes | 96GB | Not measured | Estimated |
One verdict per workload, from the DXP6800 Ultra's use-case scores and capability rows. Unknown means we cannot confirm it, not that it probably works.
The DXP6800 Ultra is a good fit for local LLM work. Ollama is vendor confirmed, but the local LLM row is limited to CPU-only inference in Docker, with no NPU.
The DXP6800 Ultra is a good fit for private RAG. Open WebUI and Ollama are vendor confirmed, though RAG support is limited to document chat through Open WebUI.
AI search is only an OK fit here. The capability is vendor confirmed as limited: photo description and OCR search work through Model Manager, while document semantic search is not stated per model.
AI photos is a good fit. The vendor confirms an AI photo album that handles faces, places, objects and duplicates, and Quick Sync on the Iris Xe graphics supports media handling.
Surveillance is a good fit. The vendor confirms that Surveillance Center supports the DXP series, with local AI detection of people, vehicles, packages and pets, though features vary by model.
Home lab use is a good fit. Docker and virtual machines are both supported on UGOS Pro, and the two 10GbE ports, six bays and two M.2 slots give room to build.
Business use is a good fit. It has six bays, two 10GbE ports, Docker and VM support, and a 3-year warranty, though the warranty figure is medium confidence. It has no ECC memory.
Creative AI is a good fit. GPU expansion is vendor confirmed through the internal PCIe Gen4 x4 slot or a Thunderbolt 4 eGPU, and hardware acceleration is limited to Quick Sync on the Iris Xe.
Out of the box the DXP6800 Ultra has 8GB; the official ceiling is 96GB of DDR5 SO-DIMM 5200. A memory upgrade is the cheapest way to make it more useful for AI.
Working backwards from 96GB: reserve about 8GB for the operating system and storage services, and at roughly 0.6GB per billion parameters at 4-bit, the largest model that fits is around 146B (Estimated).
Partly. The DXP6800 Ultra has a 1x PCIe Gen4 x4, which suits low-profile cards and network or storage upgrades. Only small single-slot GPUs fit, so expect modest gains.
For a personal library, yes. A compact embedding model and a small chat model will run, but the DXP6800 Ultra is better as the document store behind a RAG system whose model runs on stronger hardware.
Built-in networking: 2 x 10GbE. Plenty for a shared knowledge base with several users.
| Configuration | RAM | SSD | Notes |
|---|---|---|---|
| 8GB | 8GB | None | Diskless, 8GB DDR5, 128GB system SSD. |
The closest UGREEN model by bay count and AINAS Score. Only the specifications that differ are listed.
Its strongest AINAS profiles are Home Lab at 66 and Business at 65, which points to home lab owners who want one box for storage, containers, media and a few AI services.
Other AI Ready systems closest in price.
Buying guides that list it
Workloads it suits
Yes, at Q4 only, and only once the memory is upgraded. Mistral Large 2 123B needs 73GB at Q4 and the maximum is 96GB, so it fits at Q4 (Estimated). At Q8 it needs 130GB, so it does not fit (Estimated). This is a calculation, not a benchmark, and inference runs on the CPU.
Up to 96GB. It uses DDR5 SO-DIMM 5200 memory in two slots. The system ships with 8GB, so most buyers will need to add memory to use the larger models listed on this page. The memory is user-expandable.
Yes. It has an internal PCIe Gen4 x4 slot, and two Thunderbolt 4 ports allow an eGPU. The vendor confirms GPU expansion and documents dedicated GPU use in Docker. The card is not included, and the power supply wattage is not published.
No. The Intel Core 5 120U has no NPU, so no TOPS figure applies. AI work runs on the CPU, with Quick Sync on the Iris Xe graphics for hardware acceleration, which the vendor lists as limited.
Yes. UGOS Pro on the DXP6800 Ultra runs Docker and virtual machines, and Ollama and Open WebUI are both vendor confirmed there. UGREEN also documents dedicated GPU use in Docker, so a card in the PCIe Gen4 x4 slot or over Thunderbolt 4 can serve containers. Without one, inference stays on the 10-core Intel Core 5 120U, and the 8GB fitted needs raising toward 96GB before the larger models fit (Estimated).
No. The record shows ECC as not supported. It uses non-ECC DDR5 SO-DIMM 5200 memory. Buyers who need error-correcting memory for critical data should look for a system that lists ECC support.
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 10GbE built in, and the 1x PCIe Gen4 x4 can take a faster network card. The fastest port is 2 x 10GbE.
Every field records its source. Specifications were last checked on . Spot an error? Send a correction with a manufacturer link.