Skip to content

Tools / LLM GPU calculator / RTX PRO 6000 Server

What LLMs fit on an NVIDIA RTX PRO 6000 Server?

Which of 22 open LLMs fit one RTX PRO 6000 Server, and how many it takes for the rest, at FP16, FP8 and 4-bit.

Size your own scenario

Short answer

One NVIDIA RTX PRO 6000 Server (96 GB) runs 17 of the 22 open LLMs sized here at 4-bit, 15 at FP8 and 11 at FP16, for one user at an 8K context. At FP8: every model except Llama 3.1 405B, Llama 4 Scout, Llama 4 Maverick, Qwen3-235B-A22B, DeepSeek-V3 / R1, gpt-oss-120b and GLM-5.3. The largest that fits one GPU is gpt-oss-120b (116.8B parameters), at 4-bit. Within one 8-GPU server, tensor parallelism runs every model except GLM-5.3 at FP8. Llama 3.3 70B takes 2 GPUs (tensor parallel 2) at FP16 and 1 GPU at FP8.

Updated . One user, 8K context, KV cache at FP16, plus 10% headroom for activations and the runtime.

Every model on the RTX PRO 6000 Server

ModelFP16/BF16FP84-bit (INT4)
Llama 3.1 8B1 GPU19 GB1 GPU10 GB1 GPU6.1 GB
Qwen3 8B1 GPU20 GB1 GPU10 GB1 GPU6.3 GB
Qwen3 14B1 GPU34 GB1 GPU18 GB1 GPU11 GB
Phi-4 14B1 GPU35 GB1 GPU18 GB1 GPU11 GB
gpt-oss-20b1 GPU47 GB1 GPU23 GB1 GPU13 GB
LFM2.5-24B-A2B1 GPU54 GB1 GPU27 GB1 GPU15 GB
Mistral Small 3.2 24B1 GPU55 GB1 GPU28 GB1 GPU16 GB
Gemma 3 27B1 GPU62 GB1 GPU32 GB1 GPU18 GB
Qwen3-30B-A3B1 GPU69 GB1 GPU35 GB1 GPU20 GB
Qwen3 32B1 GPU75 GB1 GPU39 GB1 GPU22 GB
Qwen2.5 32B1 GPU75 GB1 GPU39 GB1 GPU22 GB
Mixtral 8x7B2 GPUs · TP2105 GB1 GPU53 GB1 GPU30 GB
Llama 3.3 70B2 GPUs · TP2160 GB1 GPU81 GB1 GPU46 GB
Qwen2.5 72B2 GPUs · TP2165 GB1 GPU84 GB1 GPU47 GB
Kolibri-12 GPUs · TP2174 GB1 GPU87 GB1 GPU48 GB
Llama 4 Scout4 GPUs · TP4243 GB2 GPUs · TP2122 GB1 GPU68 GB
gpt-oss-120b4 GPUs · TP4260 GB2 GPUs · TP2130 GB1 GPU72 GB
Qwen3-235B-A22BMulti-server524 GB4 GPUs · TP4263 GB2 GPUs · TP2145 GB
Llama 4 MaverickMulti-server894 GB8 GPUs · TP8448 GB4 GPUs · TP4247 GB
Llama 3.1 405BMulti-server907 GB8 GPUs · TP8456 GB4 GPUs · TP4253 GB
DeepSeek-V3 / R1Multi-server1,492 GB8 GPUs · TP8746 GB8 GPUs · TP8411 GB
GLM-5.3Multi-server1,654 GBMulti-server827 GB8 GPUs · TP8455 GB

Counts use a tensor-parallel size (TP) that divides the model's attention heads, within one 8-GPU server. Below each count is the memory needed: (weights + KV cache) ÷ 0.9. Capacity only: speed depends on memory bandwidth and the serving stack.

Memory is step one. Throughput, latency, power and the network decide the real build.

Size my actual workload

RTX PRO 6000 Server specifications

Memory
96 GB GDDR7
Memory bandwidth
1,597 GB/s
Typical system
PCIe server
Counted as
Tensor parallel within one 8-GPU server (vLLM style)

Questions about the RTX PRO 6000 Server

What LLMs can run on an NVIDIA RTX PRO 6000 Server?

For one user at an 8K context, these fit one RTX PRO 6000 Server. At 4-bit: every model except Llama 3.1 405B, Llama 4 Maverick, Qwen3-235B-A22B, DeepSeek-V3 / R1 and GLM-5.3. At FP8: every model except Llama 3.1 405B, Llama 4 Scout, Llama 4 Maverick, Qwen3-235B-A22B, DeepSeek-V3 / R1, gpt-oss-120b and GLM-5.3. At FP16/BF16: Llama 3.1 8B, Qwen3 8B, Qwen3 14B, Phi-4 14B, gpt-oss-20b, LFM2.5-24B-A2B, Mistral Small 3.2 24B, Gemma 3 27B, Qwen3-30B-A3B, Qwen3 32B and Qwen2.5 32B.

What is the largest LLM an RTX PRO 6000 Server can run?

Of the models sized here, for one user at an 8K context: at 4-bit, gpt-oss-120b (116.8B parameters, 72 GB); at FP8, Kolibri-1 (78.1B parameters, 87 GB); at FP16, Qwen3 32B (32.8B parameters, 75 GB). Mixture-of-experts models count every expert, because all of them must be in memory.

How many NVIDIA RTX PRO 6000 Server GPUs does Llama 3.3 70B need?

For one user at an 8K context it needs 160 GB at FP16/BF16, 81 GB at FP8 and 46 GB at 4-bit: 2 GPUs (tensor parallel 2) at FP16 and 1 GPU at FP8 or 4-bit.

Can an NVIDIA RTX PRO 6000 Server run DeepSeek-V3 / R1?

Not on one GPU: with all 671B parameters resident it needs 746 GB at FP8 and 411 GB at 4-bit for one user at an 8K context. That is 8 GPUs (tensor parallel 8) at FP8 or 4-bit.

Other GPUs

Different context length, more users, or a model that isn’t listed? The LLM GPU calculator sizes any Hugging Face model for any of these GPUs.

Need to deploy private AI?

We design it, validate it, source it, deploy it and help run it.

Get my 48-hour blueprint

Give us your workload, users, budget and timeline. Within 48 hours we'll give you a validated cloud/on-prem architecture, BOM, expected costs and sourcing options.