Can the NVIDIA L40S run MiniMax H3 Image?

Yes, it runs great (encoder in RAM)

Runs great (encoder in RAM). At FP8 it needs about 38.6 GB of 48.0 GB usable and takes about 8.7 s per image.

Best precision: FP8
Memory mode: Text encoder offloaded
Time per image: 8.7 s
Output: 2048×2048, 6 steps

Best app to run it on the NVIDIA L40S

On Linux

ComfyUI

Desktop app · runs on PyTorch (ComfyUI), GGUF via ComfyUI-GGUF

Runs this model with built-in or popular custom nodes; GGUF files load with ComfyUI-GGUF.

Get ComfyUI

On Windows

ComfyUI

Desktop app · runs on PyTorch (ComfyUI), GGUF via ComfyUI-GGUF

Runs this model with built-in or popular custom nodes; GGUF files load with ComfyUI-GGUF.

Get ComfyUI

How we choose: the files published for this model, your operating system and GPU, whether it fits fully in memory, and published speed tests (MLX vs llama.cpp on Apple silicon, vLLM vs llama.cpp vs Ollama on NVIDIA). Details.

Download MiniMax H3 Image

Original weights · Hugging Face

Full-precision files from MiniMax, for fine-tuning, vLLM, Transformers, Diffusers or ComfyUI.

huggingface.co/MiniMaxAI/MiniMax-H3-Image (opens in a new tab)

No sign-in needed.

Official repo · GitHub

Reference code, inference scripts and docs from MiniMax. The weights themselves download from Hugging Face.

github.com/MiniMax-AI/MiniMax-H3-Image (opens in a new tab)

git clone https://github.com/MiniMax-AI/MiniMax-H3-Image

Links checked 5 hours ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 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.: How capable the model itself is at FP8, from its editorial quality tier scaled by the quantization or precision. It’s the same on any hardware that runs this variant. Scale: A+ 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Runs greatencoder in RAM: How well it runs on the NVIDIA L40S: speed, capped by how well it fits in memory. Runs great: 30+ tokens/s (or near the image, video or decision speed target) · Runs well: 10+ tokens/s · Slow: 3+ tokens/s · Very slow: under 3 · Won’t run: doesn’t fit even with offload. A tight fit or RAM offload caps it at Runs well; heavy offload caps it at Slow.

Match score

To order lists, each result also gets a match score out of 100 from four parts. It favors the most capable model that still runs well.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.	41.0 / 45
Memory fitFull marks when everything sits in GPU memory; fewer when part spills to system RAM.	20.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.	15.0 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	15.0 / 15
Match score	91.0 / 100

The match score orders results; it is not shown as a grade. Quality and Runs are the two ratings in the tables.

Every precision on the NVIDIA L40S

Quant	Memory needed	Available	Speed	Memory mode	Runs	Quality	Download
`FP16`	27.5 GB	48.0 GB	12 s	Sequential CPU offload	Runs greatCPU offload	93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking.	Original weights (opens in a new tab)
`FP8`	38.6 GB	48.0 GB	8.7 s	Text encoder offloaded	Runs greatencoder in RAM	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.	Original weights (opens in a new tab)

Other hardware that runs it

91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA RTX PRO 6000 Blackwell FP8 · 6.6 s · Runs great
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. AMD Ryzen AI Max+ 395 (128 GB) FP8 · 17 s · Runs well
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. NVIDIA RTX Spark N1X (6144-core GPU) (128 GB) FP8 · 20 s · Runs well
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. NVIDIA GB10 (128 GB) FP8 · 19 s · Runs well
91Excellent quality (A+) · 91 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. Apple M2 Ultra (192 GB) FP8 · 39 s · Runs well
91Excellent quality (A+) · 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. Apple M3 Ultra (256 GB) FP8 · 38 s · Runs well

Quick answers

How much memory does MiniMax H3 Image need on the NVIDIA L40S 48GB?

At FP8 it needs about 38.6 GB. The NVIDIA L40S 48GB has 48.0 GB usable.

How long does MiniMax H3 Image take on the NVIDIA L40S 48GB?

About 8.7 s per image at FP8 and default settings, an estimate scaled from a reference timing by this hardware’s compute speed.

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.