Lenovo ThinkStation PGX GB10 128GB
128 GB unified · 273 GB/s
In short
The Lenovo ThinkStation PGX GB10 128GB has 128 GB of unified memory, of which the GPU can use 119 GB by default, with 273 GB/s of memory bandwidth. At 32k context it runs 80 of the 129 open models we track fully in memory; the most capable is DeepSeek V4 Flash at UD-Q2_K_XL, at about 11 tok/s (estimated). We have no current price for the Lenovo ThinkStation PGX GB10 128GB in Germany.
Updated . Speeds are estimates; how they are worked out.
Prices
Every retailer we track in Germany, cheapest first. Buy links may earn us a commission.
No retailer in this market lists it right now.
What it runs
Every model in the catalogue, with the best quant for this hardware. Estimates, not measurements.
| Model | Quant | Fits | Memory | Max context | Decode @32k | Prefill | Good for |
|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash DeepSeek | UD-Q2_K_XL | fits | 93.8 GB | 312k557k with q8 cache | 11tok/sest. | 210 tok/s | – |
| GLM 5.3 Flash Z.ai | UD-Q2_K_XL | fits | 102.6 GB | 1018k1024k with q8 cache | 11tok/sest. | 282 tok/s | – |
| Qwen 3.8 Flash Next Alibaba Qwen | Q4_K_M | fits | 113.1 GB | 232k256k with q8 cache | 2tok/sest. | 218 tok/s | – |
| MiniMax M2.7 MiniMax | UD-Q3_K_M | fits | 102.9 GB | 96k178k with q8 cache | 8tok/sest. | 593 tok/s | – |
| Qwen 3.8 27B Alibaba Qwen | Q8_0 | fits | 30.1 GB | 256k | 5tok/sest. | 450 tok/s | – |
| Qwen 3.5 122B-A10B Alibaba Qwen | Q6_K | fits | 95.9 GB | 256k | 11tok/sest. | 1055 tok/s | – |
| Qwen 3.6 27B Alibaba Qwen | Q8_0 | fits | 29.7 GB | 256k | 6tok/sest. | 450 tok/s | – |
| gpt-oss 120B OpenAI | MXFP4 | fits | 61.1 GB | 128k | 23tok/sest. | 1704 tok/s | chat, coding |
| Qwen3 Coder Next Alibaba Qwen | Q8_0 | fits | 80.7 GB | 256k | 20tok/sest. | 2543 tok/s | chat, coding |
| Nemotron 3 Super 120B-A12B NVIDIA | UD-Q6_K | fits | 108.2 GB | 256k | 9tok/sest. | 1067 tok/s | – |
| Gemma 4 31B | Q8_0 | fits | 34.6 GB | 256k | 5tok/sest. | 401 tok/s | – |
| Mistral Small 4 119B Mistral AI | UD-Q6_K | fits | 94.2 GB | 923k1024k with q8 cache | 15tok/sest. | 1173 tok/s | chat |
| Qwen 3.6 35B-A3B Alibaba Qwen | Q8_0 | fits | 35.9 GB | 256k | 27tok/sest. | 3213 tok/s | chat, coding, long documents |
| Qwen 3.5 27B Alibaba Qwen | Q8_0 | fits | 29.7 GB | 256k | 6tok/sest. | 450 tok/s | – |
| Gemma 4 26B-A4B | Q8_0 | fits | 26.7 GB | 256k | 20tok/sest. | 2858 tok/s | chat, coding |
| GLM 4.5 Air Z.ai | Q8_0 | fits | 116.0 GB | 48k87k with q8 cache | 5tok/sest. | 436 tok/s | – |
| Qwen 3.5 35B-A3B Alibaba Qwen | Q8_0 | fits | 35.9 GB | 256k | 27tok/sest. | 3213 tok/s | chat, coding, long documents |
| Qwen3 Next 80B-A3B Alibaba Qwen | Q8_0 | fits | 80.7 GB | 256k | 25tok/sest. | 3026 tok/s | chat, coding, long documents |
| GLM 4.7 Flash Z.ai | Q8_0 | fits | 32.2 GB | 198k | 14tok/sest. | 1677 tok/s | chat |
| Devstral Small 2 24B Mistral AI | Q8_0 | fits | 29.2 GB | 384k | 6tok/sest. | 475 tok/s | – |
| Nemotron 3.5 Lightning 30B-A3B NVIDIA | UD-Q6_K_XL | fits | 33.7 GB | 256k | 28tok/sest. | 3666 tok/s | chat, coding, long documents |
| gpt-oss 20B OpenAI | MXFP4 | fits | 12.9 GB | 128k | 31tok/sest. | 2398 tok/s | chat, coding, agents, long documents |
| Kimi Linear 48B-A3B Moonshot AI | Q8_0 | fits | 49.8 GB | 1024k | 29tok/sest. | 2769 tok/s | chat, coding, long documents |
| Qwen3 Coder 30B-A3B Alibaba Qwen | Q8_0 | fits | 34.1 GB | 256k | 16tok/sest. | 1477 tok/s | chat |
| Qwen3 32B Alibaba Qwen | Q8_0 | fits | 41.3 GB | 40k | 4tok/sest. | 279 tok/s | – |
| Gemma 4 12B | Q8_0 | fits | 13.5 GB | 256k | 13tok/sest. | 1044 tok/s | – |
| Qwen 3.5 9B Alibaba Qwen | Q8_0 | fits | 10.8 GB | 256k | 16tok/sest. | 1301 tok/s | chat |
| Mistral Small 3.2 24B Mistral AI | Q8_0 | fits | 29.2 GB | 128k | 6tok/sest. | 475 tok/s | – |
| Gemma 3 27B | Q8_0 | fits | 30.5 GB | 128k | 5tok/sest. | 478 tok/s | – |
| DeepSeek R1 Distill Qwen 32B DeepSeek | Q8_0 | fits | 41.3 GB | 128k | 4tok/sest. | 321 tok/s | – |
| Llama 3.3 70B Meta | Q8_0 | fits | 80.7 GB | 128k | 2tok/sest. | 152 tok/s | – |
| Ministral 3 14B Mistral AI | Q8_0 | fits | 19.3 GB | 256k | 9tok/sest. | 722 tok/s | – |
| Qwen3 14B Alibaba Qwen | Q8_0 | fits | 20.5 GB | 40k | 8tok/sest. | 650 tok/s | – |
| DeepSeek R1 0528 Qwen3 8B DeepSeek | Q8_0 | fits | 13.5 GB | 128k | 13tok/sest. | 1071 tok/s | chat |
| Qwen3 8B Alibaba Qwen | Q8_0 | fits | 13.5 GB | 40k | 13tok/sest. | 1071 tok/s | chat |
| Phi-4 14B Microsoft | Q8_0 | fits | 21.6 GB | 16k | 8tok/sest. | 653 tok/s | – |
| Qwen 3.5 4B Alibaba Qwen | Q8_0 | fits | 6.1 GB | 256k | 30tok/sest. | 2433 tok/s | chat, coding, long documents |
| Ministral 3 8B Mistral AI | Q8_0 | fits | 13.5 GB | 256k | 13tok/sest. | 1035 tok/s | chat |
| Granite 4.2 8B IBM | Q8_0 | fits | 14.6 GB | 128k | 12tok/sest. | 985 tok/s | chat |
| Gemma 4 E4B | Q8_0 | fits | 9.4 GB | 128k | 19tok/sest. | 1634 tok/s | chat |
| LFM2.5 8B-A1B Liquid AI | Q8_0 | fits | 10.8 GB | 125k | 38tok/sest. | 5344 tok/s | chat, coding |
| Llama 3.1 8B Meta | Q6_K | fits | 11.0 GB | 128k | 16tok/sest. | 1132 tok/s | chat |
| Qwen2.5 0.5B Alibaba Qwen | Q8_0 | fits | 1.8 GB | 32k | 181tok/sest. | 11638 tok/s | chat, coding, long documents |
| Qwen3.5 0.8B Alibaba Qwen | Q8_0 | fits | 2.0 GB | 256k | 137tok/sest. | 10932 tok/s | chat, coding, agents, long documents |
| Qwen3.5 2B Alibaba Qwen | Q5_K_M | fits | 2.6 GB | 256k | 91tok/sest. | 5212 tok/s | chat, coding, agents, long documents |
| LFM3 1.7B Liquid AI | Q8_0 | fits | 3.0 GB | 256k | 75tok/sest. | 6487 tok/s | chat, coding, agents, long documents |
| DeepSeek R1 Distill Qwen 1.5B DeepSeek | Q8_0 | fits | 3.5 GB | 128k | 60tok/sest. | 4378 tok/s | chat, coding, long documents |
| MiniCPM5 2B openbmb | Q8_0 | fits | 4.7 GB | 128k | 41tok/sest. | 2615 tok/s | chat, coding, agents, long documents |
| Spark X2.5 4B XHToken | mlx-4bit | fits | 4.9 GB | 1024k | 39tok/sest. | 2568 tok/s | chat, coding, agents, long documents |
| LFM2.5 2.6B Liquid AI | Q8_0 | fits | 5.4 GB | 128k | 35tok/sest. | 2962 tok/s | chat, coding, agents, long documents |
| LFM2.5 VL 3B Liquid AI | Q8_0 | fits | 5.4 GB | 32k | 35tok/sest. | 2716 tok/s | chat, coding, agents, long documents |
| Mellum2.1 12B A2.5B Thinking JetBrains | Q8_0 | fits | 13.4 GB | 128k | 32tok/sest. | 3858 tok/s | chat, coding, agents, long documents |
| FrogNano 4B 2609 Microsoft | Q8_0 | fits | 6.2 GB | 256k | 30tok/sest. | 2433 tok/s | chat, coding, long documents |
| granite vision 4.1 4b IBM | Q8_0 | fits | 6.8 GB | 128k | 27tok/sest. | 1897 tok/s | chat, coding |
| granite 4.2 3b IBM | Q8_0 | fits | 7.0 GB | 128k | 26tok/sest. | 1989 tok/s | chat, coding |
| Kimi VL A3B Thinking 2506 Moonshot AI | Q8_0 | fits | 17.6 GB | 128k | 25tok/sest. | 2899 tok/s | chat, coding |
| Kimi VL A3B Moonshot AI | mlx-8bit | fits | 18.4 GB | 128k | 24tok/sest. | 2440 tok/s | chat, coding |
| Ministral 3 3B Instruct 2512 Mistral AI | Q8_0 | fits | 7.5 GB | 256k | 24tok/sest. | 1901 tok/s | chat, coding |
| LFM2.5 24B A2B Liquid AI | Q8_0 | fits | 27.1 GB | 128k | 21tok/sest. | 2978 tok/s | chat, coding |
| GLM 4 9B 0414 Z.ai | mlx-5bit | fits | 9.8 GB | 32k | 18tok/sest. | 945 tok/s | chat |
| DeepSeek R1 Distill Qwen 7B DeepSeek | Q8_0 | fits | 10.2 GB | 128k | 17tok/sest. | 1279 tok/s | chat |
| GLM 4.6V Flash Z.ai | Q8_0 | fits | 11.5 GB | 128k | 15tok/sest. | 891 tok/s | chat |
| DeepSeek R1 Distill Llama 8B DeepSeek | Q8_0 | fits | 12.8 GB | 128k | 13tok/sest. | 1132 tok/s | chat |
| GLM 4.1V 9B Thinking Z.ai | mlx-8bit | fits | 13.1 GB | 64k | 13tok/sest. | 891 tok/s | – |
| Ministral 3 8B Reasoning 2512 Mistral AI | Q8_0 | fits | 13.5 GB | 256k | 13tok/sest. | 1035 tok/s | chat |
| Ministral 3 14B Reasoning 2512 Mistral AI | Q8_0 | fits | 19.3 GB | 256k | 9tok/sest. | 722 tok/s | – |
| DeepSeek R1 Distill Qwen 14B DeepSeek | Q8_0 | fits | 21.5 GB | 128k | 8tok/sest. | 611 tok/s | – |
| Qwen3.8 27B OBLITERATED OBLITERATUS | mlx-4bit | fits | 22.3 GB | 256k | 7tok/sest. | 450 tok/s | – |
| Devstral Small 2507 Mistral AI | Q8_0 | fits | 29.2 GB | 128k | 6tok/sest. | 482 tok/s | – |
| Magistral Small 2506 Mistral AI | Q8_0 | fits | 29.2 GB | 40k | 6tok/sest. | 482 tok/s | – |
| Magistral Small 2507 Mistral AI | Q8_0 | fits | 29.2 GB | 40k | 6tok/sest. | 482 tok/s | – |
| Magistral Small 2509 Mistral AI | Q8_0 | fits | 29.2 GB | 128k | 6tok/sest. | 475 tok/s | – |
| Muse Glimmer 30B meta-models | Q8_0 | fits | 31.5 GB | 128k | 5tok/sest. | 434 tok/s | – |
| GLM 4.5V Z.ai | Q8_0 | fits | 112.4 GB | 64k | 5tok/sest. | 425 tok/s | – |
| GLM 4.6V Z.ai | Q8_0 | fits | 112.4 GB | 67k123k with q8 cache | 5tok/sest. | 425 tok/s | – |
| granite 4.2 30b IBM | Q8_0 | fits | 37.9 GB | 128k | 4tok/sest. | 368 tok/s | – |
| DeepSeek R1 Distill Llama 70B DeepSeek | Q8_0 | fits | 80.7 GB | 128k | 2tok/sest. | 152 tok/s | – |
| Kimi Dev 72B Moonshot AI | Q8_0 | fits | 82.8 GB | 128k | 2tok/sest. | 148 tok/s | – |
| Devstral 2 123B Instruct 2512 Mistral AI | Q6_K | fits | 107.4 GB | 65k120k with q8 cache | 2tok/sest. | 87 tok/s | – |
| Mistral Medium 3.5 128B Mistral AI | Q6_K | fits | 107.4 GB | 65k120k with q8 cache | 2tok/sest. | 86 tok/s | – |
| DeepSeek R1 DeepSeek | Q4_K_M | won't run | 379.7 GB | – | – | – | – |
| DeepSeek R1 0528 DeepSeek | Q4_K_M | won't run | 380.2 GB | – | – | – | – |
| DeepSeek R1 Zero DeepSeek | Q4_K_M | won't run | 379.7 GB | – | – | – | – |
| DeepSeek V3 DeepSeek | Q4_K_M | won't run | 379.7 GB | – | – | – | – |
| DeepSeek V3 0324 DeepSeek | Q4_K_M | won't run | 379.7 GB | – | – | – | – |
| DeepSeek V3.1 DeepSeek | Q4_K_M | won't run | 380.6 GB | – | – | – | – |
| DeepSeek V3.1 Terminus DeepSeek | Q4_K_M | won't run | 380.6 GB | – | – | – | – |
| DeepSeek V3.2 DeepSeek | Q4_K_M | won't run | 380.6 GB | – | – | – | – |
| DeepSeek V4 Flash 0731 DeepSeek | MXFP4 | won't run | 149.2 GB | – | – | – | – |
| DeepSeek V4 Flash Vision Exp DeepSeek | MXFP4 | won't run | 158.0 GB | – | – | – | – |
| GLM 4.5 Z.ai | Q4_K_M | won't run | 214.0 GB | – | – | – | – |
| GLM 4.6 Z.ai | Q4_K_M | won't run | 213.2 GB | – | – | – | – |
| GLM 4.7 Z.ai | Q4_K_M | won't run | 214.0 GB | – | – | – | – |
| GLM 5 Z.ai | MXFP4 | won't run | 386.0 GB | – | – | – | – |
| GLM 5.1 Z.ai | MXFP4 | won't run | 423.6 GB | – | – | – | – |
| GLM 5.2 Z.ai | UD-Q4_K_M | won't run | 437.5 GB | – | – | – | – |
| GLM 5.3 Z.ai | Q8_0 | won't run | 750.0 GB | – | – | – | – |
| Kimi K2 Moonshot AI | Q4_K_M | won't run | 581.2 GB | – | – | – | – |
| Kimi K2 Instruct 0905 Moonshot AI | Q4_K_M | won't run | 581.6 GB | – | – | – | – |
| Kimi K2 Thinking Moonshot AI | Q4_K_M | won't run | 581.6 GB | – | – | – | – |
| Kimi K2.5 Moonshot AI | Q4_K_M | won't run | 581.6 GB | – | – | – | – |
| MiniMax M2 MiniMax | Q4_K_M | won't run | 137.5 GB | – | – | – | – |
| MiniMax M2.1 MiniMax | Q4_K_M | won't run | 137.5 GB | – | – | – | – |
| MiniMax M2.5 MiniMax | MXFP4 | won't run | 123.9 GB | 12k22k with q8 cache | – | – | – |
| MiniMax M3 MiniMax | MXFP4 | won't run | 244.3 GB | – | – | – | – |
Specifications
From the manufacturer's published specifications; sources are linked where we have them.
Compare with
Questions
- How much memory does the Lenovo ThinkStation PGX GB10 128GB have for AI models?
- The Lenovo ThinkStation PGX GB10 128GB has 128 GB of unified memory, of which the GPU can use 119 GB by default, with 273 GB/s of memory bandwidth.
- What is the best model the Lenovo ThinkStation PGX GB10 128GB can run locally?
- DeepSeek V4 Flash (UD-Q2_K_XL) fits fully in its memory at 32k context and generates at about 11 tok/s, an estimate from its memory bandwidth. 80 of the 129 models we track fit at that context.
- Can the Lenovo ThinkStation PGX GB10 128GB run DeepSeek R1?
- No: even at Q4_K_M, DeepSeek R1 needs about 380 GB at 32k context, more than the Lenovo ThinkStation PGX GB10 128GB and system RAM can hold.
- How much does the Lenovo ThinkStation PGX GB10 128GB cost in Germany?
- We have no current price for the Lenovo ThinkStation PGX GB10 128GB in Germany.
Speeds on this page are for 32k context. See the best hardware for local AI at every budget.