What can a RTX 5090 32GB run?

32GB · 1792 GB/s · 575W · $2400

The enthusiast sinkhole that finally fits 70B-class Q3/Q4 and makes 8B scream at 150+ tok/s.

Runs great30

Comfortable headroom — daily-driver territory.

Model	Quant	Est. speed	Max context
deepseek moe 16b chat	FP16	Infinity tok/s	—
qwen2 7b instruct	FP16	Infinity tok/s	—
Qwen3-0.6B	FP16	895 tok/s	33k
Qwen3-30B-A3B	Q5_K_M	472 tok/s	33k
Qwen3-Coder-30B-A3B	Q5_K_M	472 tok/s	49k
Gemma-3n-E4B	FP16	396 tok/s	131k
SmolLM2-1.7B	FP16	326 tok/s	8k
GPT-OSS-20B	Q8_0	290 tok/s	50k
Llama-3.2-3B	FP16	173 tok/s	131k
Phi-4-mini	FP16	144 tok/s	131k
Gemma-3-4B	FP16	144 tok/s	131k
Qwen3-4B	FP16	142 tok/s	33k
Gemma-3-12B	Q8_0	81 tok/s	122k
Mistral-7B-Instruct	FP16	77 tok/s	33k
DeepSeek-R1-Distill-Qwen-7B	FP16	73 tok/s	126k
Qwen2.5-Coder-7B	FP16	73 tok/s	126k
Phi-4-14B	Q8_0	71 tok/s	16k
DeepSeek-R1-Distill-Qwen-14B	Q8_0	71 tok/s	104k
Llama-3.1-8B	FP16	69 tok/s	119k
Qwen3-8B	FP16	69 tok/s	33k
Qwen3-14B	Q8_0	62 tok/s	33k
Codestral-22B	Q6_K	61 tok/s	33k
Qwen3-32B	Q4_K_M	56 tok/s	33k
DeepSeek-R1-Distill-Qwen-32B	Q4_K_M	56 tok/s	61k
GLM-4-32B	Q4_K_M	56 tok/s	61k
Mistral-Small-24B	Q6_K	56 tok/s	64k
Qwen3.8	Q6_K	50 tok/s	47k
Qwen3.8 27B	Q6_K	50 tok/s	47k
Gemma-3-27B	Q6_K	50 tok/s	47k
qwen35 27B	Q6_K	49 tok/s	45k

Tight squeeze2

It fits, but close the browser tabs and mind the context.

Model	Quant	Est. speed	Max context
Llama-3.3-70B	IQ2_M	47 tok/s	31k
Qwen2.5-72B	IQ2_M	45 tok/s	27k

Measured on this card0

Everything here was benchmarked on this exact card by someone who owns one — the tables above are our arithmetic; these numbers came off real machines.

Nobody has measured anything on this card yet. Paste a benchmark — it takes one keystroke and no form.

Try your own machine →← All hardware