VOOZH about

URL: https://willitrunai.com/can-run/llava-1.6-13b-on-rtx-pro-6000-blackwell-server-96gb

⇱ LLaVA 1.6 13B on RTX PRO 6000 Blackwell Server Edition 96GB…

Can LLaVA 1.6 13B run on RTX PRO 6000 Blackwell Server Edition 96GB?

YES — Runs Great

A72Great
○Estimated from fit model

LLaVA 1.6 13B needs ~30.9 GB VRAM. RTX PRO 6000 Blackwell Server Edition 96GB has 96.0 GB. With Q4_K_M quantization, expect ~169 tok/s.

Runtime: OllamaCapacity: RoomyBandwidth: HighStack: BasicBottleneck: Balanced
Share:

Operating mode

Choose the run profile you care about

Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.

Current mode

Balanced

Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 30.9 GB, 169.2 tok/s, Runs well
30.9 GB required96.0 GB available
32% VRAM used

Fit status

Runs well

Decode

169.2 tok/s

TTFT

1144 ms

Safe context

4K

Memory

30.9 GB / 96.0 GB

Memory breakdown

Weights7.9 GB
KV Cache12.2 GB
Runtime1.2 GB
Headroom9.6 GB

See how fast it feels

See how fast it feelsLLaVA 1.6 13B on RTX PRO 6000 Blackwell Server Edition 96GB
1st promptCold start — includes initialization
>What is local AI inference?
Local AI inference means running an AI language model directly on your own hardware — your laptop, desktop, or server — instead of sending requests to a remote cloud API. When you run inference locally the model weights are loaded into your GPU or unified memory. Each token you generate requires reading those weights from memory, so memory bandwidth is the main bottleneck for decode speed. Key benefits of running locally: - Full privacy: your prompts never leave your machine - No per-token cost or rate limits - Works offline once the model is downloaded - Latency depends only on your hardware
2nd promptWarm — model ready, no init overhead
>How much VRAM do I need?
It depends on the model size and quantization level. A rough rule of thumb: Model size Q4 (4-bit) Q8 (8-bit) FP16 7B params ~4.3 GB ~7.5 GB ~14 GB 13B params ~7.9 GB ~13.9 GB ~26 GB 70B params ~42.7 GB ~74.9 GB ~140 GB Most people use 4-bit quantization (Q4_K_M) which gives 90-95% of full quality at a fraction of the memory. A 24 GB GPU can comfortably run most 7B-13B models.
Estimated: 169.2 tok/s decode · 1.1s TTFT (warm) · 423 tok/s prefill

What limits this setup

This setup is broadly balanced for this model.

No major red flags

This recommendation has enough memory headroom and acceptable estimated speed for the selected workload.

Best improvement path

Performance by workload

WorkloadGradeFitDecodeTTFTContext
ChatARuns well169.2 tok/s624 ms4K
CodingARuns well169.2 tok/s1144 ms4K
Agentic CodingARuns well169.2 tok/s1665 ms4K
ReasoningARuns well169.2 tok/s1353 ms4K
RAGARuns well169.2 tok/s2081 ms4K

Quantization options

How LLaVA 1.6 13B (13B params) fits at each quantization level on RTX PRO 6000 Blackwell Server Edition 96GB (96.0 GB usable).

QuantBitsVRAMQualityFit
Q2_K
2
5.1 GB
LowB62
Q3_K_S
3
6.4 GB
LowB62
NVFP4
4
7.3 GB
MediumB63
Q4_K_M
4
7.9 GB
MediumB63
Q5_K_M
5
9.4 GB
HighB63
Q6_K
6
10.7 GB
HighB63
Q8_0
8
13.9 GB
Very HighB63
F16Best for your GPU
16
26.7 GB
MaximumB65

Get started

Copy-paste commands to run LLaVA 1.6 13B on your machine.

Run

docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \ --hf-repo "liuhaotian/llava-v1.6-mistral-7b" \ --hf-file "llava-v1.6-mistral-7b-Q4_K_M.gguf" \ -c 4096 -ngl 99

Your hardware

More models your RTX PRO 6000 Blackwell Server Edition 96GB can run

ModelParamsGradeDecodeCapabilities
👁 Mistral
Devstral 2 123B Instruct		123BS19.4 tok/s
👁 Alibaba
Qwen3-Coder 30B A3B Instruct		30.5BS202.8 tok/s
👁 Alibaba
Qwen 3.5 27B		27BS88 tok/s
👁 Alibaba
Qwen 3.6 27B		27BS88.2 tok/s
👁 Alibaba
Qwen 3.5 122B A10B		122BS53.9 tok/s

Frequently asked questions

See all results for RTX PRO 6000 Blackwell Server Edition 96GBSee all hardware for LLaVA 1.6 13B