Can Mistral Small 4 119B run on NVIDIA DGX Spark 128GB?

YES — Tight Fit

S86Excellent

Estimated from fit model

Mistral Small 4 119B needs ~91.9 GB VRAM. NVIDIA DGX Spark 128GB has 108.8 GB. With Q4_K_M quantization, expect ~7 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) — 91.9 GB, 7.1 tok/s, Tight fit

91.9 GB required108.8 GB available

84% VRAM used

Fit status

Tight fit

Decode

7.1 tok/s

TTFT

27115 ms

Safe context

66K

Memory

91.9 GB / 108.8 GB

Memory breakdown

Weights72.6 GB

KV Cache5.4 GB

Runtime0.9 GB

Headroom13.1 GB

See how fast it feels

See how fast it feelsMistral Small 4 119B on NVIDIA DGX Spark 128GB
1st promptCold start — includes initialization
>Write a Python function to sort a list
Here's a Python function that sorts a list using quicksort:

def quicksort(arr):
    if len(arr) <= 1:
        return arr
    pivot = arr[len(arr) // 2]
    left   = [x for x in arr if x < pivot]
    middle = [x for x in arr if x == pivot]
    right  = [x for x in arr if x > pivot]
    return quicksort(left) + middle + quicksort(right)

# Example usage
numbers = [3, 6, 8, 10, 1, 2, 1]
print(quicksort(numbers))  # [1, 1, 2, 3, 6, 8, 10]

The algorithm picks a pivot, partitions the list into smaller, equal, and larger elements, then recursively sorts each partition and concatenates the results.
2nd promptWarm — model ready, no init overhead
>Now add type hints and a docstring
def quicksort(arr: list[int]) -> list[int]:
    """Sort a list using the quicksort algorithm.

    Args:
        arr: List of integers to sort.

    Returns:
        A new sorted list.
    """
    if len(arr) <= 1:
        return arr
    pivot = arr[len(arr) // 2]
    left   = [x for x in arr if x < pivot]
    middle = [x for x in arr if x == pivot]
    right  = [x for x in arr if x > pivot]
    return quicksort(left) + middle + quicksort(right)

Added type hints for the input and return type, plus a Google-style docstring.
Estimated: 7.1 tok/s decode · 27.1s TTFT (warm) · 18 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	S	Tight fit	7.1 tok/s	14790 ms	66K
Coding	S	Tight fit	7.1 tok/s	27115 ms	66K
Agentic Coding	S	Tight fit	7.1 tok/s	39441 ms	66K
Reasoning	S	Tight fit	7.1 tok/s	32045 ms	66K
RAG	S	Tight fit	7.1 tok/s	49301 ms	66K

Inference speed

Mistral Small 4 119B inference speed — tokens per second by GPU & Mac

Estimated decode speed (tokens/sec) for Mistral Small 4 119B at Q4_K_M across popular GPUs and Apple Silicon, including multi-GPU rigs, using the fastest local runtime per device. Fastest is Mac Studio M3 Ultra 256GB at ~38 tok/s. Speed is memory-bandwidth bound, so cards that fit the whole model in VRAM run far faster than ones that offload to system RAM.

GPU / Mac	Memory	Quant	Speed (tok/s)	Fits?
Mac Studio M3 Ultra 256GB	256 GB	Q4_K_M	37.6	Fits
Mac Studio M2 Ultra 128GB	128 GB	Q4_K_M	30.8	Offloads
Mac Studio M1 Ultra 128GB	128 GB	Q4_K_M	29.3	Offloads
MacBook Pro M4 Max 128GB	128 GB	Q4_K_M	22.9	Offloads
2× RX 7900 XTX 24GB	48 GB	Q4_K_M	11.9	Too big
MacBook Pro M4 Max 64GB	64 GB	Q4_K_M	10.7	Too big
2× RTX 4090 24GB	48 GB	Q4_K_M	8.0	Too big
RTX 5090 32GB	32 GB	Q4_K_M	7.9	Too big
MacBook Pro M3 Max 64GB	64 GB	Q4_K_M	7.5	Too big
MacBook Pro M1 Max 64GB	64 GB	Q4_K_M	6.8	Too big
2× RTX 3090 24GB	48 GB	Q4_K_M	6.8	Too big
MacBook Pro M4 Pro 48GB	48 GB	Q4_K_M	6.4	Too big
4× RTX 3060 12GB	48 GB	Q4_K_M	6.0	Too big
RTX 4090 24GB	24 GB	Q4_K_M	5.0	Too big
RX 7900 XTX 24GB	24 GB	Q4_K_M	4.5	Too big
RTX 3090 24GB	24 GB	Q4_K_M	4.3	Too big
RTX 4080 Super 16GB	16 GB	Q4_K_M	4.0	Too big
RTX 4070 12GB	12 GB	Q4_K_M	2.5	Too big
RTX 3060 12GB	12 GB	Q4_K_M	2.0	Too big
RTX 4060 8GB	8 GB	Q4_K_M	2.0	Too big

Estimates for single-stream decoding at Q4_K_M; real tokens/sec varies with prompt length, context, batch size, and runtime build. Prompt processing (prefill) is faster than the decode figures shown here.

Quantization options

How Mistral Small 4 119B (119B params) fits at each quantization level on NVIDIA DGX Spark 128GB (92.2 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	46.4 GB	Low	S88
Q3_K_S	3	58.3 GB	Low	S88
NVFP4	4	66.6 GB	Medium	S88
Q4_K_MBest for your GPU	4	72.6 GB	Medium	S88
Q5_K_M	5	85.7 GB	High	F0
Q6_K	6	97.6 GB	High	F0
Q8_0	8	127.3 GB	Very High	F0
F16	16	244.0 GB	Maximum	F0

Get started

Copy-paste commands to run Mistral Small 4 119B on your machine.

Run

lms load Mistral-Small-4-119B-2603 && lms server start

Your hardware

More models your NVIDIA DGX Spark 128GB can run

Model	Params	Grade	Decode	Capabilities
Devstral 2 123B Instruct	123B	S	2.4 tok/s
Qwen 3.5 122B A10B	122B	S	6.6 tok/s

Frequently asked questions

Can NVIDIA DGX Spark 128GB run Mistral Small 4 119B?

Yes, NVIDIA DGX Spark 128GB can run Mistral Small 4 119B with a S grade (Tight fit). Expected decode speed: 7.1 tok/s.

How much VRAM does Mistral Small 4 119B need?

Mistral Small 4 119B (119B parameters) requires approximately 91.9 GB of memory with Q4_K_M quantization.

What is the best quantization for Mistral Small 4 119B?

The recommended quantization for Mistral Small 4 119B is Q4_K_M, which balances quality and memory efficiency.

What speed will Mistral Small 4 119B run at on NVIDIA DGX Spark 128GB?

On NVIDIA DGX Spark 128GB, Mistral Small 4 119B achieves approximately 7.1 tokens per second decode speed with a time-to-first-token of 27115ms using Q4_K_M quantization.

Can NVIDIA DGX Spark 128GB run Mistral Small 4 119B for coding?

For coding workloads, Mistral Small 4 119B on NVIDIA DGX Spark 128GB receives a S grade with 7.1 tok/s and 66K context.

What context window can Mistral Small 4 119B use on NVIDIA DGX Spark 128GB?

On NVIDIA DGX Spark 128GB, Mistral Small 4 119B can safely use up to 66K tokens of context. The model's official context limit is 256K, but available memory constrains the safe maximum.

What should I upgrade first if Mistral Small 4 119B feels slow on NVIDIA DGX Spark 128GB?

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.

Is unified memory on NVIDIA DGX Spark 128GB as fast as VRAM for Mistral Small 4 119B?

Not always. NVIDIA DGX Spark 128GB can often fit larger models thanks to unified memory, but a discrete GPU with dedicated high-bandwidth VRAM may still decode faster once the model fits. For this combination, the important distinction is capacity versus sustained throughput.

See all results for NVIDIA DGX Spark 128GB See all hardware for Mistral Small 4 119B

Embed this result▼

Paste this snippet into any page to show a live fit card.

<iframe src="https://willitrunai.com/embed/mistral-small-4-119b-on-dgx-spark-128gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>

Preview: