Can Gemma 4 26B A4B run on NVIDIA L40S 48GB?

YES — Runs Great

S88Excellent

▲Estimated — low-sample bucket· few comparable runs

Gemma 4 26B A4B needs ~24.7 GB VRAM. NVIDIA L40S 48GB has 48.0 GB. With Q4_K_M quantization, expect ~99 tok/s.

Runtime: llama.cppCapacity: RoomyBandwidth: HighStack: StandardBottleneck: Balanced

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) — 24.7 GB, 98.5 tok/s, Runs well

24.7 GB required48.0 GB available

51% VRAM used

Fit status

Runs well

Decode

98.5 tok/s

TTFT

1966 ms

Safe context

118K

Memory

24.7 GB / 48.0 GB

Memory breakdown

Weights15.4 GB

KV Cache3.7 GB

Runtime0.9 GB

Headroom4.8 GB

See how fast it feels

See how fast it feelsGemma 4 26B A4B on NVIDIA L40S 48GB
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: 98.5 tok/s decode · 2.0s TTFT (warm) · 246 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

Workload	Grade	Fit	Decode	TTFT	Context
Chat	S	Runs well	98.5 tok/s	1072 ms	118K
Coding	S	Runs well	98.5 tok/s	1966 ms	118K
Agentic Coding	S	Runs well	98.5 tok/s	2859 ms	118K
Reasoning	S	Runs well	98.5 tok/s	2323 ms	118K
RAG	S	Runs well	98.5 tok/s	3574 ms	118K

Quantization options

How Gemma 4 26B A4B (25.200000762939453B params) fits at each quantization level on NVIDIA L40S 48GB (48.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	9.8 GB	Low	A78
Q3_K_S	3	12.3 GB	Low	A79
NVFP4	4	14.1 GB	Medium	A79
Q4_K_M	4	15.4 GB	Medium	A80
Q5_K_M	5	18.1 GB	High	A80
Q6_K	6	20.7 GB	High	A81
Q8_0Best for your GPU	8	27.0 GB	Very High	A83
F16	16	51.7 GB	Maximum	F0

Get started

Copy-paste commands to run Gemma 4 26B A4B on your machine.

Run

ollama run gemma4:26b

Your hardware

More models your NVIDIA L40S 48GB can run

Model	Params	Grade	Decode
👁 Alibaba Qwen3-Coder 30B A3B Instruct	30.5B	S	73.4 tok/s
👁 Alibaba Qwen 3.5 27B	27B	S	30.6 tok/s
👁 Alibaba Qwen 3.6 27B	27B	S	20.1 tok/s
👁 Alibaba Qwen 3.6 35B A3B	35B	S	91.6 tok/s
👁 Alibaba Qwen3-VL 30B A3B Instruct	30B	S	105.4 tok/s

Frequently asked questions

See all results for NVIDIA L40S 48GB See all hardware for Gemma 4 26B A4B

URL: https://willitrunai.com/can-run/gemma-4-26b-a4b-on-l40s-48gb

⇱ Gemma 4 26B A4B on NVIDIA L40S 48GB? YES