# NVIDIA RTX PRO 6000 Blackwell for local LLMs

> The most memory on a single PCIe card at RTX 5090-class bandwidth, at a very steep price.

The RTX PRO 6000 has 96GB of GDDR7 ECC and 1,792 GB/s of memory bandwidth, which decides how big a model fits and how fast it generates text. It comfortably runs models up to about 122B parameters: the largest in our catalog is Qwen3.5 122B-A10B at Q4_K_M (77GB with 8K context, ~105 tokens/sec). Our top all-round pick for it is Qwen3.8 27B (Q8_0, ~46 tokens/sec).

## Key specs

- Memory: 96GB GDDR7 ECC
- Memory bandwidth: 1,792 GB/s
- Architecture: Blackwell (2025)
- Power: 600W
- Typical price: ~$16,000 (checked 2026-10)
- Software: CUDA, Vulkan

## Best models for this GPU

- **Best overall (also coding, reasoning, vision):** Qwen3.8 27B at Q8_0, ~31.6GB with 8K context, ~46 tokens/sec
- **Runner-up (also largest that fits):** Qwen3.5 122B-A10B at Q4_K_M, ~77.0GB with 8K context, ~105 tokens/sec
- **Fastest good model:** gpt-oss-20b at MXFP4, ~15.0GB with 8K context, ~190 tokens/sec

## Compatibility (best quantization that fits, 8K context)

| Model | Params | Quant | Memory | Speed | Verdict |
| --- | --- | --- | --- | --- | --- |
| Gemma 4 E4B | 8B | Q8_0 | 10.5GB | ~130 tok/s | Runs well |
| Qwen3.5 9B | 9B | Q8_0 | 12.0GB | ~115 tok/s | Runs well |
| Phi-4 | 14.7B | Q8_0 | 18.3GB | ~84 tok/s | Runs well |
| gpt-oss-20b | 20.9B (3.6B active) | MXFP4 | 15.0GB | ~190 tok/s | Runs well |
| Gemma 4 26B A4B | 26B (3.8B active) | Q8_0 | 29.5GB | ~140 tok/s | Runs well |
| Qwen3.8 27B | 27B | Q8_0 | 31.6GB | ~46 tok/s | Runs well |
| Gemma 4 31B | 31B | Q8_0 | 35.9GB | ~42 tok/s | Runs well |
| Qwen3.6 35B-A3B | 35B (3B active) | Q8_0 | 41.0GB | ~155 tok/s | Runs well |
| GLM-4.5-Air | 106B (12B active) | Q5_K_M | 79.9GB | ~80 tok/s | Runs well |
| Llama 4 Scout | 109B (17B active) | Q5_K_M | 81.3GB | ~61 tok/s | Runs well |
| gpt-oss-120b | 117B (5.1B active) | MXFP4 | 68.3GB | ~170 tok/s | Runs well |
| Qwen3.5 122B-A10B | 122B (10B active) | Q4_K_M | 77.0GB | ~105 tok/s | Runs well |
| DeepSeek V4 Flash | 284B (13B active) | - | - | - | Won't fit |
| GLM-5.3-Flash | 320B (18B active) | - | - | - | Won't fit |
| GLM-5.3 | 744B (40B active) | - | - | - | Won't fit |

## Pros
- 96GB on one card avoids multi-GPU complexity
- 1,792 GB/s matches the RTX 5090
- 300W Max-Q version fits four to a workstation
- ECC memory and professional drivers

## Cons
- Around $16,000 in October 2026
- The 600W Workstation Edition needs serious cooling

Full page: https://llamabuilds.ai/gpus/rtx-pro-6000-blackwell · Best models: https://llamabuilds.ai/gpus/rtx-pro-6000-blackwell/best-models

Sources: [NVIDIA RTX PRO 6000 Blackwell Server Edition product page](https://www.nvidia.com/en-us/data-center/rtx-pro-6000-blackwell-server-edition/), [Wikipedia, Blackwell (microarchitecture)](https://en.wikipedia.org/wiki/Blackwell_(microarchitecture))

---
Source: https://llamabuilds.ai/gpus/rtx-pro-6000-blackwell · Updated October 14, 2026 · Llama Builds. Speeds marked ~ are estimates from our bandwidth model (https://llamabuilds.ai/methodology); measured results are labeled as such.