NVIDIA · Ada Lovelace

Best AI Models for NVIDIA RTX 5000 Ada Generation (32 GB)

Runs 27–31B and 35B MoE models well; 70B doesn't fit in 32 GB.

32 GB GDDR6576 GB/s261 TFLOPS FP16250 WAda Lovelace 2023

What LLMs can the RTX 5000 Ada Generation run?

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 18.4 t/s	18.4 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 111 t/s	111 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 149 t/s	149 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 18.4 t/s	18.4 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 101 t/s	101 tok/s	Q4_K_M · 22.0 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 111 t/s	111 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 149 t/s	149 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 18.4 t/s	18.4 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.5 27BAlibaba Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s	Q4_K_M · 17.9 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 111 t/s	111 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 149 t/s	149 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Math Score	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s		Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s		Q4_K_M · 17.9 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 101 t/s	101 tok/s		Q4_K_M · 22.0 GB
4	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 18.4 t/s	18.4 tok/s	overall	Q4_K_M · 21.7 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 149 t/s	149 tok/s		Q4_K_M · 19.3 GB

#	Model	Speed ↓	Quant · VRAM
1	Qwen3 30B A3B Thinking 2507AlibabaHow to run → Q4_K_M · 19.4 GB · 123 t/s	123 tok/s	Q4_K_M · 19.4 GB
2	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 149 t/s	149 tok/s	Q4_K_M · 19.3 GB
3	Qwen3 30B A3B Instruct 2507Alibaba Q4_K_M · 19.4 GB · 123 t/s	123 tok/s	Q4_K_M · 19.4 GB
4	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 121 t/s	121 tok/s	Q4_K_M · 13.1 GB
5	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 111 t/s	111 tok/s	Q4_K_M · 16.5 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 21.5 t/s	21.5 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 18.4 t/s	18.4 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 111 t/s	111 tok/s	Q4_K_M · 16.5 GB
5	MiniCPM V 4.7 35B A3BOpenBMB Q4_K_M · 22.1 GB · 114 t/s	114 tok/s	Q4_K_M · 22.1 GB

See all 346 models on the RTX 5000 Ada Generation →Best quant that fits · speeds estimated on this card

About the RTX 5000 Ada Generation

The RTX 5000 Ada Generation runs 27–31B and 35B MoE models well. Ada Lovelace generation, released 2023.

GenerationAda Lovelacereleased 2023

RTX 5000 Ada Generation strengths and weaknesses

Strengths

Runs 35B MoE models: Qwen3.6 35B A3B at 101 tok/s in 22 GB
Room for long context: Qwen3.8 27B at Q4 uses 17.9 GB, leaving 14.1 GB free
Reads long prompts fast: A 4K-token prompt on Qwen3.8 27B starts answering in 1.9 s

Weaknesses

70B models don't fit: Llama 3.3 70B Instruct needs 51.3 GB at Q4 vs 32 GB here

RTX 5000 Ada Generation tokens per second by model size

Output speed depends on memory bandwidth (576 GB/s). Grey rows need more than 32 GB.

9BQwen3.5 9B · Q4_K_M · 6.6 GB
58.9tok/s
14BQwen3 14B · Q4_K_M · 11.9 GB
39.4tok/s
27–31BQwen3.8 27B · Q4_K_M · 17.9 GB
21.5tok/s
35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
101tok/s
70BLlama 3.3 70B Instruct · not at 4-bit
needs 51.3 GB
120B MoEGPT OSS 120B · not at 4-bit
needs 70.8 GB

Estimated speeds: Q4_K_M, 4K context, one user. How we estimate

All 1176 models on the RTX 5000 Ada Generation

Featured only

Best quant that fits, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works

LLM models compatible with NVIDIA RTX 5000 Ada Generation — ranked by Score, recency and Fit
Model	Quant	VRAM	Speed	Context	Fit
Qwen3.5 9B Abliterated9.0BChatFunctions Q4_K_M·63.0 t/s	Q4_K_M	6.2 GB19%	63.0 t/s	262K	A84, Runs well
Apertus 8B MeditronFO8.1BChat Q4_K_M·69.3 t/s	Q4_K_M	7.3 GB23%	69.3 t/s	66K	S86, Runs great
Qwen3.5 9B Uncensored9BChat Q4_K_M·63.0 t/s	Q4_K_M	8.6 GB27%	63.0 t/s	—	A84, Runs well
Phi Tiny MoE Instruct3.8BChat Q4_K_M·230.1 t/s	Q4_K_M	3.1 GB10%	230.1 t/s	4K	S100, Runs great
Small Qwen3ForCausalLM20MChatFunctions Q4_K_M·1207.7 t/s	Q4_K_M	0.3 GB1%	1207.7 t/s	41K	S100, Runs great
Falcon H1 7B Instruct7.6BChatFunctions Q4_K_M·74.3 t/s	Q4_K_M	6.3 GB20%	74.3 t/s	262K	S86, Runs great
Aria25.3BVision Q4_K_M·108.0 t/s	Q4_K_M	20.2 GB63%	108.0 t/s	66K	S93, Runs great
Olmo Hybrid 7B7BChat BF16·24.5 t/s	BF16	22.4 GB70%	24.5 t/s	66K	A65, Runs well
Tiny Random LlamaForCausalLM1MChat Q4_K_M·1248.0 t/s	Q4_K_M	0.3 GB1%	1248.0 t/s	8K	S100, Runs great
Huihui Qwen3 Coder 30B A3B Instruct Abliterated30.5BChatCodeFunctions Q4_K_M·123.0 t/s	Q4_K_M	19.4 GB61%	123.0 t/s	262K	S95, Runs great
Qwen3 4B Instruct 2507 Heretic4.0BChat Q4_K_M·129.1 t/s	Q4_K_M	4.2 GB13%	129.1 t/s	262K	S95, Runs great
JOSIE 1.1 4B Instruct4.0BChat Q4_K_M·129.1 t/s	Q4_K_M	4.2 GB13%	129.1 t/s	262K	S95, Runs great
Qwen3 8B8BChat Q4_K_M·70.9 t/s	Q4_K_M	7.8 GB25%	70.9 t/s	—	S86, Runs great
Granite 4.1 3B3.4BChatFunctions Q4_K_M·149.2 t/s	Q4_K_M	3.7 GB12%	149.2 t/s	131K	S97, Runs great
Protgpt2 Distilled Tiny39MChat Q4_K_M·12480.0 t/s	Q4_K_M	0.0 GB0%	12480.0 t/s	1K	S100, Runs great

Rent NVIDIA RTX 5000 Ada Generation in the cloud

Hourly rate on demand for one RTX 5000 Ada Generation.

Vast.aiMarketplace
from$0.44/hr
Deploy

Buying one ($4,000) costs as much as ~9,000 hours of rental, about 38 months at 8 hours a day.

Cost per million tokens

Output-token cost on an RTX 5000 Ada Generation rented at $0.44/hr, one request at a time. Batch several users on the same card and the cost per token drops several-fold.

Model	Quant	Speed	Per 1M output
NVIDIA Nemotron 3.5 Lightning 30B A3B BF16	Q4_K_M	149 tok/s	$0.83$0.0049 input
Gemma 4 26B A4B IT	Q4_K_M	111 tok/s	$1.11$0.0092 input
Qwen3.6 35B A3B	Q4_K_M	101 tok/s	$1.22$0.0099 input
Qwen3.8 27B	Q4_K_M	22 tok/s	$5.72$0.058 input
Muse Glimmer 30B	Q4_K_M	20 tok/s	$6.03$0.062 input

RTX 5000 Ada Generation specs, power and PSU

Specs

Architecture: Ada Lovelace
VRAM: 32 GB GDDR6
Memory bandwidth: 576.0 GB/s
FP16 (dense): 261.10 TFLOPS
CUDA cores: 12,800
Tensor cores: 400
Compute capability: 8.9
TDP: 250 W
Released: 2023-08-09
MSRP: $4,000

RTX 5000 Ada Generation FAQ

What is the largest model the RTX 5000 Ada Generation can run?

Qwen3.6 35B A3B (36.0B) is the largest at 4-bit or better: at Q4_K_M it uses 22.0 GB of 32 GB, and runs at about 101 tok/s.

Can the RTX 5000 Ada Generation run Qwen3.8 27B?

Yes. Qwen3.8 27B fits at Q4_K_M, using 17.9 GB of 32 GB, and it is the best model that runs well on this card.

Can the RTX 5000 Ada Generation run 70B models?

No. Llama 3.3 70B Instruct needs 51.3 GB at Q4, more than the 32 GB on this card.

How many tokens per second does the RTX 5000 Ada Generation give?

About 21.5 tok/s on Qwen3.8 27B at Q4_K_M, with the first token after 1.9 s on a 4K-token prompt.

What power supply does the RTX 5000 Ada Generation need?

Plan a 500 W PSU (2× the 250 W TDP).

Is 32 GB enough for coding and agents?

Yes for coding: Qwen3.8 27B, the best of the 10 coding models that fit, runs at about 22 tok/s. For agents, not yet: no agent-scored model fits in less than 96 GB.

Best coding LLMs 24 GB VRAM 48 GB VRAM GPUs with 32 GB or more RTX 5090