What AI models can the NVIDIA L40S run?

VRAM: 48.0 GB
Usable for models: 48.0 GB
Memory bandwidth: 864 GB/s
FP16 compute: 91.6 TFLOPS
Released: 2023

523 / 622 LLMs run Best: MiniCPM-V 4.7 35B-A3B 61 / 61 Image models run Best: Krea 2 Turbo 42 / 43 Video models run Best: LTX-2.5 (22B, video + audio) 69 / 70 Decision models run Best: Qwen3 Reranker 4B

More about this hardware

Maker	NVIDIA
Type	Graphics card
Interface	PCIe 4.0

Apps that run models well on the NVIDIA L40S

Linux

vLLMThe full model fits in GPU memory at 8-bit, so it can batch many requests: best for serving a team or an app.
OllamaOne command with Ollama’s own build of this model.
LM StudioPoint-and-click: search, download and chat in one window.
SGLangThe full model fits in GPU memory at 8-bit, so it can batch many requests: best for serving a team or an app.

Windows

For a typical model that fits in memory. Each model page picks the best app for that model on this hardware. All apps

1 product with the NVIDIA L40S

TensorEX Rackmount Workstation - TWS-154715021Exxact · Workstation · 4, 8, 20, 24, 32, 40, 48, 94 or 96 GBProduct page (opens in a new tab)

Best LLMs for the NVIDIA L40S

Quality	Model	Quant	Memory	Speed	Runs
93Excellent quality (A+) · 93 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking.	MiniCPM-V 4.7 35B-A3BOpenBMB	`Q8_0`	40.5 GB	107 tok/s	Runs great
92Excellent quality (A+) · 92 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking.	Qwen3.6 35B-A3BAlibaba	`Q6_K`	32.6 GB	130 tok/s	Runs great
92Excellent quality (A+) · 92 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking.	Qwen3.5 35B-A3BAlibaba	`Q6_K`	32.6 GB	130 tok/s	Runs great
92Excellent quality (A+) · 92 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking.	Gemma 4 26B-A4BGoogle	`Q8_0`	30.8 GB	92.2 tok/s	Runs great
92Excellent quality (A+) · 92 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking.	Nex N2 Mini (35B-A3B)Nex-AGI	`Q8_0`	40.4 GB	107 tok/s	Runs great
91Excellent quality (A+) · 91 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking.	KAT-Coder V2.5 Dev (35B-A3B)Kuaishou	`Q8_0`	39.9 GB	107 tok/s	Runs great
90Excellent quality (A+) · 90 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking.	Sage-A1.5 35B-A3BInternScience	`Q6_K`	32.6 GB	130 tok/s	Runs great
89Excellent quality (A+) · 89 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking.	Flint (40B-A2.5B)Recursive	`Q6_K`	35.9 GB	143 tok/s	Runs great
89Excellent quality (A+) · 89 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking.	Ornith 1.5 35B-A3BDeepReinforce	`Q6_K`	32.6 GB	130 tok/s	Runs great
88Excellent quality (A+) · 88 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking.	Apodex 1.1 Mini (35B-A3B)Apodex	`Q6_K`	32.6 GB	130 tok/s	Runs great
88Excellent quality (A+) · 88 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking.	Ornith 1.0 35B-A3BDeepReinforce	`Q6_K`	32.6 GB	130 tok/s	Runs great
88Excellent quality (A+) · 88 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking.	GLM-4.7 Flash (30B-A3B)Zhipu AI	`Q8_0`	36.5 GB	107 tok/s	Runs great

Best Image models for the NVIDIA L40S

Quality	Model	Quant	Memory	Speed	Runs
92Excellent quality (A+) · 92 of 100: how capable the model is at FP16, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking.	Krea 2 TurboKrea	`FP16`	37.4 GB	4.0 s	Runs great
87Excellent quality (A+) · 87 of 100: how capable the model is at FP8, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking.	Qwen-Image 2.1 TurboAlibaba	`FP8`	30.8 GB	9.7 s	Runs great
84Great quality (A) · 84 of 100: how capable the model is at FP16, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking.	FLUX.2 [klein] 9BBlack Forest Labs	`FP16`	37.1 GB	1.7 s	Runs great
91Excellent quality (A+) · 91 of 100: how capable the model is at FP16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking.	FLUX 3 Image KleinBlack Forest Labs	`FP16`	22.4 GB	5.3 s	Runs great
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking.	MiniMax H3 ImageMiniMax	`FP8`	38.6 GB	8.7 s	Runs greatencoder in RAM
85Excellent quality (A+) · 85 of 100: how capable the model is at FP16, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking.	ERNIE-Image TurboBaidu	`FP16`	26.4 GB	3.0 s	Runs great

Best Video models for the NVIDIA L40S

Quality	Model	Quant	Memory	Speed	Runs
81Great quality (A) · 81 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking.	LTX-2.5 (22B, video + audio)Lightricks	`Q4 GGUF`	39.9 GB	1.7 min	Runs great
84Great quality (A) · 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking.	LTX-2.3Lightricks	`FP8`	44.3 GB	5.6 min	Runs welltight fit
80Great quality (A) · 80 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking.	LTX-2 (19B, video + audio)Lightricks	`FP8`	44.3 GB	5.6 min	Runs welltight fit
66Good quality (B) · 66 of 100: how capable the model is at FP8, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking.	LTX-Video 13B 0.9.8 (distilled)Lightricks	`FP8`	30.1 GB	50 s	Runs great
84Great quality (A) · 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking.	Wan 2.2 I2V A14BAlibaba	`FP8`	38.3 GB	22.2 min	Slow

Best Decision models for the NVIDIA L40S

Quality	Model	Quant	Memory	Speed	Runs
90Excellent quality (A+) · 90 of 100: how capable the model is at F16, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking.	Qwen3 Reranker 4BAlibaba	`F16`	11.3 GB	51 ms	Runs great
90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking.	Qwen3 Reranker 8BAlibaba	`Q8_0`	12.0 GB	83 ms	Runs great
88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking.	zerank-2ZeroEntropy	`F16`	11.3 GB	51 ms	Runs great
86Excellent quality (A+) · 86 of 100: how capable the model is at F16, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking.	Jina Reranker v3.5Jina AI	`F16`	4.4 GB	11 ms	Runs great
86Excellent quality (A+) · 86 of 100: how capable the model is at F16, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking.	Shieldstral 1.0 3BMistral AI	`F16`	10.9 GB	49 ms	Runs great
90Excellent quality (A+) · 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking.	Skywork Reward V2 Llama 3.1 8BSkywork	`Q8_0`	10.4 GB	131 ms	Runs great
84Great quality (A) · 84 of 100: how capable the model is at F16, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking.	Intern-Decision 4BShanghai AI Lab	`F16`	12.1 GB	40 ms	Runs great
84Great quality (A) · 84 of 100: how capable the model is at F16, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking.	Jina Reranker v3Jina AI	`F16`	4.4 GB	11 ms	Runs great

Best LLMs by task on this hardware Open in Model Finder

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.