Products · QNAP · NAS
QNAP TVS-AIh1688ATX
- Processor
- Intel Core Ultra 5 225 / 7 265 / 9 285 (desktop)
- Memory (RAM)
- 32, 48 or 192 GB
- Released
- 2025
Configurations
Pick one to see which models it runs. More memory lets larger models fit; bandwidth sets how fast they answer.
| Configuration | Memory | Usable for AI | Bandwidth | LLMs that run well | Launch price |
|---|---|---|---|---|---|
| Intel Core Ultra 5 225 / 7 265 / 9 285 (desktop) | 32.0 GB | 26.2 GB | 90 GB/s | 381 of 622 | — |
| Intel Core Ultra 5 225 / 7 265 / 9 285 (desktop) | 48.0 GB | 40.3 GB | 90 GB/s | 387 of 622 | — |
| Intel Core Ultra 5 225 / 7 265 / 9 285 (desktop) · upgraded to the maker's maximum | 192 GB | 173 GB | 90 GB/s | 417 of 622 | — |
Specifications 192 GB configuration
| Maker | QNAP |
|---|---|
| Type | NAS |
| Processor | Intel Core Ultra 5 225 / 7 265 / 9 285 (desktop) NAS (192 GB) · upgraded to the maker's maximum |
| Memory (RAM) | 192 GB |
| Usable for AI models | 173 GB |
| Memory bandwidth | 90 GB/s |
| FP16 compute | 0.8 TFLOPS |
| Runs models on | CPU cores |
| Drive bays | 12 x 3.5-inch SATA 6Gb/s, 3Gb/s + 4 x 2.5-inch U.2 NVMe Gen 4 x2 slots |
| NAS system | QuTS hero |
| Containers (Docker) | Yes |
| Virtual machines | Yes |
| Memory upgrade | DDR5 UDIMM · 4 slots · up to 192 GB per the maker · ECC |
| Graphics card options | PCIe: 3 ; Slot 1: PCIe Gen4 x16 ; Slot 2: PCIe Gen4 x4 ; Slot 3: PCIe Gen4 x4. External GPU: 2 x USB4 ports; optional Thunderbolt 5 via QXP-T52P card. eGPU use not documented by QNAP. Maker-supported graphics cards: QNAP FAQ: graphics card in PCIe slot 1 (Gen4 x16). NVIDIA driver packages cover GTX 10 to RTX 50 / RTX PRO; Intel Arc B70 support added in QuTS hero h6.1. |
| Notes | Intel Core Ultra 9 24-core (8P+16E)/24-thread, up to 5.6 GHz or Intel Core Ultra 7 20-core (8P+12E)/20-thread, up to…. Core Ultra 7/9 with NPU and iGPU, 32-48 GB ECC DDR5 (max 192 GB), and a PCIe Gen4 x16 slot for a full discrete GPU. Runs Ollama in Container Station on CPU; with an NVIDIA (or, from h6.1, Intel Arc B70) card it runs QNAP's on-prem Qsirch LLM and Qsirch AI Mode. |
| Product page | www.qnap.com (opens in a new tab) |
Running AI on the QNAP TVS-AIh1688ATX
Yes, it runs AI models on the NAS itself. Models run on its CPU cores, and this configuration leaves about 173 GB for them after the NAS system and its storage services. The grades below assume the NAS isn’t busy with backups or media work at the same time. More memory is the upgrade that matters most: see the configurations.
Apps you can install to run models
| App | How it installs |
|---|---|
| Ollama + Open WebUI | In a container (Docker) |
| Ollama / Open WebUI or any other LLM server image | In a container (Docker) |
| Any Linux VM running an LLM runtime | In a virtual machine |
| Ollama | In a container (Docker) |
| Open WebUI | In a container (Docker) |
AI features from QNAP
| Feature | What it does | Where the AI runs |
|---|---|---|
| Qsirch RAG Search (cloud LLM mode) | Natural-language search and Q&A over NAS files; retrieval runs on the NAS but the LLM is a cloud API (Gemini, OpenAI, Azure OpenAI or OpenAI-compatible). | Cloud service or another computer |
| QNAP AI Core (QuMagie, QVR Face, QVR Human, Qsirch OCR) | On-device image-recognition engine (faces, objects, similar photos, people counting, OCR) accelerated by the NPU / Intel CPU+GPU+NPU. Vision models only; the product page names no on-device LLM for it. | On the NAS |
| QuMagie + QNAP AI Core | Photo management with on-device face, object and scene recognition (QNAP AI Core image-recognition engine). | On the NAS |
| Qsirch AI Semantic Search | Natural-language/semantic file and image search computed locally; QNAP requires a 64-bit x86 NAS with at least 8 GB RAM. | On the NAS |
| Qsirch on-premises RAG / LLM Core | Runs open models (DeepSeek R1, Gemma 3, gpt-oss 20B, Mistral 7B, Phi-4) locally for RAG search. Requires an NVIDIA GPU assigned to Container Station; QNAP states CPU-only is not supported. Needs QuTS hero h5.2.1+ and Qsirch 6.0.0+. | On the NAS |
| QuAgent (beta) + AI Service Center (beta) | Conversational NAS assistant on QTS 6.1 / QuTS hero h6.1; uses either a cloud model or an on-premises model (compatible GPU recommended). QNAP lists x86 NAS "x74 series and above", TVS-AI and QAI series as supported. | On the NAS |
Best LLMs for the QNAP TVS-AIh1688ATX 192 GB
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 93Excellent quality (A+) · 93 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | Qwen3.8 Flash Next (125B-A6B) | Q5_K_M | 142 GB | 12.1 tok/s | Runs well |
| 86Excellent quality (A+) · 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | Kolibri 1 (78B-A3.5B) | Q4_K_M | 62.2 GB | 24.3 tok/s | Runs well |
| 84Great quality (A) · 84 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | gpt-oss 120B | MXFP4 | 81.3 GB | 18.1 tok/s | Runs well |
| 78Great quality (A) · 78 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | MiniMax M2.7 (230B-A10B) | Q3_K_M | 129 GB | 11.4 tok/s | Runs well |
| 85Excellent quality (A+) · 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | Solar Open 2 250B | Q4_K_M | 169 GB | 6.0 tok/s | Slowtight fit |
| 78Great quality (A) · 78 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | Step 3.7 Flash (198B-A11B) | Q3_K_M | 114 GB | 10.4 tok/s | Runs well |
| 77Great quality (A) · 77 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | MiniMax M2.5 (230B-A10B) | Q3_K_M | 129 GB | 11.4 tok/s | Runs well |
| 81Great quality (A) · 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | Qwen3-Next 80B-A3B Thinking | Q4_K_M | 64.2 GB | 27.7 tok/s | Runs well |
| 86Excellent quality (A+) · 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | Qwen3 235B-A22B Thinking 2507 | Q4_K_M | 159 GB | 4.1 tok/s | Slowtight fit |
| 81Great quality (A) · 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | Qwen3 Coder Next (80B-A3B) | Q4_K_M | 63.2 GB | 27.7 tok/s | Runs well |
Best Image models for the QNAP TVS-AIh1688ATX 192 GB
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | MiniMax H3 Image | FP16 | 138 GB | 9.7 min | Very slow |
| 92Excellent quality (A+) · 92 of 100: how capable the model is at FP16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | FLUX.2 [dev] | FP16 | 116 GB | 35.0 min | Very slow |
| 92Excellent quality (A+) · 92 of 100: how capable the model is at FP16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | Cosmos3-Super Text2Image (64B) | FP16 | 133 GB | 38.9 min | Very slow |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. | HunyuanImage 3.0 (80B-A13B) | FP8 | 86.9 GB | 52.5 min | Very slow |
| 92Excellent quality (A+) · 92 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | Ideogram 4 | FP16 | 57.3 GB | 24.3 min | Very slow |
| 93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | Qwen-Image 2.1 | FP16 | 57.0 GB | 58.3 min | Very slow |
Best Video models for the QNAP TVS-AIh1688ATX 192 GB
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 90Excellent quality (A+) · 90 of 100: how capable the model is at FP16, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | MiniMax H3 (33B, video + audio) | FP16 | 113 GB | 16.8 h | Very slow |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | Cosmos3-Super Image2Video (64B) | FP16 | 139 GB | 36.4 h | Very slow |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | LTX-2.5 (22B, video + audio) | FP16 | 82.5 GB | 88.4 min | Very slow |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | Kandinsky 6.0 Pro (29B, video + audio) | FP16 | 85.4 GB | 47.6 h | Very slow |
| 86Excellent quality (A+) · 86 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | LTX-2.3 | FP16 | 75.4 GB | 6.5 h | Very slow |
Best Decision models for the QNAP TVS-AIh1688ATX 192 GB
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 78Great quality (A) · 78 of 100: how capable the model is at F16, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. | OpenAI Privacy Filter | F16 | 3.8 GB | 37 ms | Runs great |
| 68Good quality (B) · 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | Llama Prompt Guard 2 22M | F16 | 0.8 GB | 99 ms | Runs great |
| 80Great quality (A) · 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | GLiNER2.5 Decide | Q8_0 | 1.2 GB | 288 ms | Runs well |
| 86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | Jina Reranker v3.5 | Q8_0 | 15.3 GB | 657 ms | Slow |
| 66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | GLiClass Base v3 | Q8_0 | 0.8 GB | 114 ms | Runs great |
| 84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | Jina Reranker v3 | Q8_0 | 15.3 GB | 657 ms | Slow |
| 64Good quality (B) · 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | DeBERTa v3 Base Zero-shot v2.0 | Q8_0 | 0.8 GB | 108 ms | Runs great |
| 74Great quality (A) · 74 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | GLiClass ModernBERT Large v3 | Q8_0 | 1.1 GB | 236 ms | Runs well |
Spec sources
Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.