How much VRAM does Qwen3 8B need?

At Q4_K_M the weights are 4.7 GB. With an 8K context the KV cache adds 1.1 GB and the runtime adds 0.8 GB, for 6.7 GB in total.

Qwen on NVIDIA Blackwell

Can I run Qwen3 8B on an RTX 5080?

Yes. Qwen3 8B at Q4_K_M uses 6.7 GB of the 14.2 GB available on a RTX 5080, and runs at about 117 tokens per second. You can push the context to 62K.

Model

Hardware

Quantisation

Context 8K

KV cache precision

Runtime

Batch

GPUs

Monitor plugged into this GPU

Runs comfortably

6.7 GB of 14.2 GB · 47%

014 GB

Weights 4.7 GB

KV cache 1.1 GB

Runtime overhead 0.8 GB

Qwen3 8B at Q4_K_M leaves 7.6 GB spare on a RTX 5080. There is room to raise the context length or move up a quantisation level.

Generation117tok/s

Prompt processing2872tok/s

Max context62Ktokens

KV per 1K tokens0GB

Highlighted row is the highest quality that still fits at 8K context.

Quantisation	bpw	Weights	Total	Verdict	Max ctx	tok/s
FP16 / BF16	16.00	15.3 GB	17.2 GB	Won't fit	—	10.7
INT8 / W8A8	8.50	7.8 GB	9.8 GB	Runs comfortably	40K	74.5
Q8_0 (GGUF)	8.50	7.8 GB	9.8 GB	Runs comfortably	40K	74.5
FP8 (E4M3)	8.00	7.6 GB	9.6 GB	Runs comfortably	41K	76.2
Q6_K	6.56	6.3 GB	8.2 GB	Runs comfortably	51K	91.2
AWQ 4-bit	4.25	5.8 GB	7.7 GB	Runs comfortably	55K	98.3
GPTQ 4-bit	4.25	5.8 GB	7.7 GB	Runs comfortably	55K	98.3
MXFP4	4.25	5.8 GB	7.7 GB	Runs comfortably	55K	98.3
Q5_K_M	5.67	5.4 GB	7.4 GB	Runs comfortably	57K	104
Q5_K_S	5.52	5.3 GB	7.2 GB	Runs comfortably	58K	106
Q4_K_M	4.85	4.7 GB	6.7 GB	Runs comfortably	62K	117
Q4_K_S	4.58	4.5 GB	6.4 GB	Runs comfortably	63K	122
Q4_0	4.55	4.5 GB	6.4 GB	Runs comfortably	64K	123
IQ4_XS	4.25	4.2 GB	6.2 GB	Runs comfortably	65K	129
Q3_K_M	3.91	4.0 GB	5.9 GB	Runs comfortably	67K	136
IQ3_M	3.70	3.8 GB	5.7 GB	Runs comfortably	68K	141
IQ3_XXS	3.06	3.3 GB	5.2 GB	Runs comfortably	72K	160
Q2_K	2.63	2.9 GB	4.9 GB	Runs comfortably	75K	175
IQ2_XXS	2.06	2.5 GB	4.4 GB	Runs comfortably	78K	201
IQ1_M	1.75	2.2 GB	4.2 GB	Runs comfortably	80K	218