Hardware

RTX 5000 Ada

RTX 5000 Ada has 32 GB GDDR6 and 576 GB/s. See which open models fit, at which quant, and roughly how fast. Figures below use the default memory (32 GB), 32 GB system RAM, and DDR5-5600 for anything that spills. Updated 2026-10-24.

Memory32 GB GDDR6

Bandwidth576 GB/s

Kindgpu

Year2023

32GB GDDR6, 576 GB/s. Source.

RSS for new models on this device

Models that fit

Model	Verdict	Quant	Context	tok/s
Qwen2.5 32B Instruct 32.8B	Runs	Q5_K_M	8K	12–17
Qwen2.5 Coder 32B Instruct 32.8B	Runs	Q5_K_M	8K	12–17
Qwen3 32B 32.8B	Tight fit	UD-Q6_K_XL	4K	11–14
Qwen3 VL 32B Instruct 32.8B	Tight fit	UD-Q6_K_XL	4K	11–14
DeepSeek R1 Distill Qwen 32B 32.8B	Tight fit	Q6_K	8K	11–15
GLM 4 32B 0414 32.6B	Tight fit	UD-Q6_K_XL	8K	11–15
Qwen3 Coder Next 79.7B	Tight fit	UD-IQ3_S	8K	54–74
DeepSeek R1 Distill Llama 70B 70.6B	Tight fit	UD-IQ3_XXS	8K	10–14
Llama 3.3 70B Instruct 70.6B	Tight fit	UD-IQ3_XXS	8K	10–14
Hermes 4 70B 70.6B	Tight fit	UD-IQ3_XXS	8K	10–14
Qwen3.8 27B 27.3B	Tight fit	UD-Q8_K_L	8K	11–15
Swift Qwen3.8 27B 27.3B	Tight fit	Q8_0	8K	11–15
gemma 3 27b it 27.0B	Tight fit	Q8_0	8K	9.6–13
medgemma 27b text it 27.0B	Tight fit	Q8_0	8K	9.6–13
Qwen3.6 27B 26.9B	Tight fit	Q8_0	8K	11–15
Qwen3.5 27B 26.9B	Tight fit	Q8_0	8K	11–15
Kwaipilot KAT Coder V2.5 Dev 34.7B	Runs	Q5_K_M	8K	67–92
Qwen3.8 35B A3B Distill 35.5B	Tight fit	Q6_K	8K	120–164
Cyber Tiel Coder 35B A3B GGUF MTP 35.5B	Tight fit	Q6_K (est.)	8K	121–164
Tiel Coder 35B A3B 34.7B	Tight fit	UD-Q5_K_XL	8K	128–175
Qwen3.5 35B A3B 34.7B	Tight fit	UD-Q6_K	8K	120–164
Qwen3.6 35B A3B 34.7B	Tight fit	UD-Q6_K	8K	117–160
NVIDIA Nemotron 3.5 Lightning 30B A3B 32.9B	Tight fit	UD-Q5_K_XL	8K	99–135
Devstral Small 2 24B Instruct 2512 23.6B	Tight fit	UD-Q8_K_XL	8K	10–14
Mistral Small 3.2 24B Instruct 2506 23.6B	Tight fit	UD-Q8_K_XL	8K	10–14
Magistral Small 2509 23.6B	Tight fit	UD-Q8_K_XL	8K	10–14
Nemotron 3 Nano 30B A3B 31.6B	Tight fit	UD-Q5_K_XL	8K	104–142
NVIDIA Nemotron 3 Nano Omni 30B A3B Reasoning 31.6B	Tight fit	UD-Q5_K_XL	8K	99–135
GLM 4.7 Flash 29.9B	Runs	UD-Q6_K_XL	8K	53–72
Qwen3 Coder 30B A3B Instruct 30.5B	Tight fit	UD-Q6_K_XL	8K	93–127
Qwen3 30B A3B Instruct 2507 30.5B	Tight fit	UD-Q6_K_XL	8K	93–127
Qwen3 VL 30B A3B Instruct 30.5B	Tight fit	UD-Q6_K_XL	8K	93–127
Qwen3 30B A3B Thinking 2507 30.5B	Tight fit	UD-Q6_K_XL	8K	93–127
Qwen2.5 72B Instruct 72.7B	Tight fit	IQ2_XXS	8K	11–15
Qwen3 Next 80B A3B Instruct 79.7B	Tight fit	UD-Q2_K_XL	4K	206–281
Qwen3 Next 80B A3B Thinking 79.7B	Tight fit	UD-Q2_K_XL	4K	206–281
bottlecapai ThinkingCap gemma 4 26B A4B it 25.2B	Tight fit	Q8_0	8K	51–69
gemma 4 26B A4B it 25.2B	Tight fit	UD-Q8_K_XL	8K	50–68
LFM2.5 24B A2B 24.0B	Runs	Q8_0	8K	113–154
GLM 4.7 Flash REAP 23B A3B 23.0B	Tight fit	UD-Q8_K_XL	8K	79–107

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

Example: Runs fully on your RTX 5000 Ada at Q5_K_M, 8K context, about 12–17 tok/s. Weights 21.7 GB plus 2.00 GB of KV cache — 24.2 GB of 29.9 GB usable. Up to 16K context still fits. Weights shown in the verdict use GiB (the same 1024³ counting as GPU memory).