Can Pixtral Large 124B run on NVIDIA DGX Spark 128GB?

YES — Tight Fit

A82Great

○Estimated from fit model

Pixtral Large 124B needs ~95.0 GB VRAM. NVIDIA DGX Spark 128GB has 108.8 GB. With Q4_K_M quantization, expect ~2 tok/s.

Runtime: llama.cppCapacity: TightBandwidth: LowStack: StandardBottleneck: Memory bandwidth

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) — 95.0 GB, 2.4 tok/s, Tight fit

95.0 GB required108.8 GB available

87% VRAM used

Fit status

Tight fit

Decode

2.4 tok/s

TTFT

82208 ms

Safe context

57K

Memory

95.0 GB / 108.8 GB

Memory breakdown

Weights75.6 GB

KV Cache5.4 GB

Runtime0.9 GB

Headroom13.1 GB

See how fast it feels

See how fast it feelsPixtral Large 124B on NVIDIA DGX Spark 128GB
1st promptCold start — includes initialization
>Solve: If a train travels at 60 mph for 2.5 hours, how far does it go?
Let me work through this step by step.

Given information:
 - Speed = 60 mph (miles per hour)
 - Time = 2.5 hours

Formula: Distance = Speed x Time

Calculation:
 Distance = 60 mph x 2.5 h
 = 150 miles

The train travels 150 miles.

Double-check: at 60 mph the train covers 60 miles each hour. In half an hour it covers 30 miles. So in 2.5 hours: 60 + 60 + 30 = 150 miles. Confirmed.
2nd promptWarm — model ready, no init overhead
>What if the train stops for 20 minutes halfway?
The stop doesn't change the total distance covered while moving, but it affects total trip time.

Breakdown:
 - First half: 1.25 hours at 60 mph = 75 miles
 - Stop: 20 minutes (no distance)
 - Second half: 1.25 hours at 60 mph = 75 miles

Total distance: 150 miles (same)
Total time: 2.5 h + 0.33 h = 2.83 hours
Average speed: 150 / 2.83 = 53 mph

The distance stays the same but average speed drops to 53 mph because of the stop.
Estimated: 2.4 tok/s decode · 82.2s TTFT (warm) · 6 tok/s prefill

What limits this setup

The model fits in shared memory, but shared-memory bandwidth is now the real limiter.

Fit does not mean dedicated-VRAM speed

Unified or shared memory can make a model technically fit, but sustained tokens per second may still trail a discrete high-bandwidth GPU with less total memory.

Shared-memory contention still exists

The OS, browser, and inference runtime all compete for the same physical memory pool, so real-world headroom is less forgiving than raw capacity suggests.

Best improvement path

Prioritize bandwidth, not only capacity

If this workload feels slow, the next useful step is often a GPU tier with materially faster memory bandwidth rather than only a small bump in capacity.

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	A	Tight fit	2.4 tok/s	44841 ms	57K
Coding	A	Tight fit	2.4 tok/s	82208 ms	57K
Agentic Coding	A	Tight fit	2.4 tok/s	119575 ms	57K
Reasoning	A	Tight fit	2.4 tok/s	97155 ms	57K
RAG	A	Tight fit	2.4 tok/s	149469 ms	57K

Quantization options

How Pixtral Large 124B (124B params) fits at each quantization level on NVIDIA DGX Spark 128GB (92.2 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	48.4 GB	Low	S87
Q3_K_S	3	60.8 GB	Low	S87
NVFP4Best for your GPU	4	69.4 GB	Medium	S87
Q4_K_M	4	75.6 GB	Medium	F0
Q5_K_M	5	89.3 GB	High	F0
Q6_K	6	101.7 GB	High	F0
Q8_0	8	132.7 GB	Very High	F0
F16	16	254.2 GB	Maximum	F0

Get started

Copy-paste commands to run Pixtral Large 124B on your machine.

Run

lms load Pixtral-Large-Instruct-2411 && lms server start

Frequently asked questions

See all results for NVIDIA DGX Spark 128GB See all hardware for Pixtral Large 124B

URL: https://willitrunai.com/can-run/pixtral-large-124b-on-dgx-spark-128gb

⇱ Pixtral Large 124B on NVIDIA DGX Spark 128GB? TIGHT FIT