Hardware

NVIDIA H100 PCIe 80GB

NVIDIA H100 PCIe 80GB has 80 GB HBM2e and 2000 GB/s. See which open models fit, at which quant, and roughly how fast. Figures below use the default memory (80 GB), 32 GB system RAM, and DDR5-5600 for anything that spills. Updated 2026-10-17.

Memory80 GB HBM2e

Bandwidth2000 GB/s

Kindgpu

Year2023

PCIe 80GB, 2 TB/s. Source.

RSS for new models on this device

Models that fit

Model	Verdict	Quant	Context	tok/s
Mistral Medium 3.5 128B 125B	Tight fit	UD-Q4_K_XL	8K	14–19
Devstral 2 123B Instruct 2512 125B	Tight fit	UD-Q4_K_XL	8K	14–19
gpt oss 120b 117B	Runs	UD-Q8_K_XL	8K	90–122
Qwen3.5 122B A10B MTP 125B	Tight fit	UD-Q4_K_XL	4K	172–234
Mistral Small 4 119B 2603 119B	Tight fit	UD-Q4_K_XL	8K	81–110
Llama 4 Scout 17B 16E Instruct 108B	Tight fit	Q5_K_M	8K	62–84
GLM 4.6V 107B	Tight fit	Q5_K_S	8K	62–85
GLM 4.5 Air 110B	Tight fit	UD-Q4_K_XL	8K	69–94
Qwen2.5 72B Instruct 72.7B	Tight fit	Q8_0	4K	14–19
Llama 3.3 70B Instruct 70.6B	Tight fit	Q8_0	8K	14–19
DeepSeek R1 Distill Llama 70B 70.6B	Tight fit	Q8_0	8K	14–19
Hermes 4 70B 70.6B	Tight fit	Q8_0	8K	14–19
NVIDIA Nemotron 3 Super 120B A12B 121B	Runs	IQ4_NL	8K	154–210
Qwen3 Coder Next 79.7B	Tight fit	UD-Q6_K_XL	8K	83–114
Qwen3 Next 80B A3B Instruct 79.7B	Tight fit	UD-Q6_K_XL	8K	325–443
Qwen3 Next 80B A3B Thinking 79.7B	Tight fit	UD-Q6_K_XL	8K	325–443
Step 3.7 Flash 197B	Tight fit	UD-Q2_K_XL	8K	46–63
MiniMax M2 229B	Tight fit	UD-IQ2_XXS	8K	71–97
Qwen3.8 Flash Next 177B	Tight fit	UD-IQ1_M	8K	86–117
Qwen2.5 Coder 32B Instruct 32.8B	Runs great	Q8_0	8K	30–41
Qwen2.5 32B Instruct 32.8B	Runs	F16	8K	16–22
Qwen3 32B 32.8B	Runs	BF16	8K	16–22
DeepSeek R1 Distill Qwen 32B 32.8B	Runs	F16	8K	16–22
Qwen3 VL 32B Instruct 32.8B	Runs	BF16	8K	16–22
GLM 4 32B 0414 32.6B	Runs	BF16	8K	17–23
gemma 4 31B it 30.7B	Runs	BF16	8K	16–22
Qwen3.8 27B 27.3B	Runs	BF16	8K	20–27
gemma 3 27b it 27.0B	Runs	BF16	8K	19–26
medgemma 27b text it 27.0B	Runs	BF16	8K	19–26
Qwen3.6 27B 26.9B	Runs	BF16	8K	20–28
Qwen3.5 27B 26.9B	Runs	BF16	8K	20–28
Kwaipilot KAT Coder V2.5 Dev 34.7B	Tight fit	BF16	8K	86–118
Qwen3.8 35B A3B Distill 35.5B	Tight fit	BF16	8K	178–243
Qwen3.6 35B A3B 34.7B	Tight fit	BF16	8K	178–243
Qwen3.5 35B A3B 34.7B	Tight fit	BF16	8K	178–243
NVIDIA Nemotron 3.5 Lightning 30B A3B 32.9B	Runs	BF16	8K	171–233
Devstral Small 2 24B Instruct 2512 23.6B	Runs	BF16	8K	23–31
Mistral Small 3.2 24B Instruct 2506 23.6B	Runs	BF16	8K	23–31
Magistral Small 2509 23.6B	Runs	BF16	8K	23–31
NVIDIA Nemotron 3 Nano Omni 30B A3B Reasoning 31.6B	Runs great	UD-Q8_K_XL	8K	290–395

90 more models in the catalog do not fit this default profile, or ranked below the first 40 that do.

Example: Tight fit fully on your NVIDIA H100 PCIe 80GB at UD-Q4_K_XL, 8K context, about 14–19 tok/s. Weights 70.5 GB plus 2.75 GB of KV cache — 74.2 GB of 76.0 GB usable. Weights shown in the verdict use GiB (the same 1024³ counting as GPU memory).