AMD · RDNA4

What runs on a Radeon RX 9070 XT

16 GB at 645 GB/s. The largest model that fits at Q4_K_M is gpt-oss 20B-A3.6B, at roughly 128 tokens per second.

Model

Hardware

Quantisation

Context 8K

KV cache precision

Runtime

Batch

GPUs

Monitor plugged into this GPU

Runs comfortably

6.4 GB of 14.2 GB · 45%

014 GB

Weights 4.6 GB

KV cache 1.0 GB

Runtime overhead 0.8 GB

Llama 3.1 8B at Q4_K_M leaves 7.8 GB spare on a Radeon RX 9070 XT. There is room to raise the context length or move up a quantisation level.

Generation74.0tok/s

Prompt processing2537tok/s

Max context70Ktokens

KV per 1K tokens0GB

Every quantisation of Llama 3.1 8B on a Radeon RX 9070 XT

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	15.0 GB	16.8 GB	Won't fit	—	9.16
INT8 / W8A8	8.50	7.7 GB	9.5 GB	Runs comfortably	46K	46.4
Q8_0 (GGUF)	8.50	7.7 GB	9.5 GB	Runs comfortably	46K	46.4
FP8 (E4M3)	8.00	7.5 GB	9.3 GB	Runs comfortably	48K	47.7
Q6_K	6.56	6.1 GB	8.0 GB	Runs comfortably	58K	57.3
AWQ 4-bit	4.25	5.4 GB	7.2 GB	Runs comfortably	64K	64.2
GPTQ 4-bit	4.25	5.4 GB	7.2 GB	Runs comfortably	64K	64.2
MXFP4	4.25	5.4 GB	7.2 GB	Runs comfortably	64K	64.2
Q5_K_M	5.67	5.3 GB	7.1 GB	Runs comfortably	65K	65.4
Q5_K_S	5.52	5.2 GB	7.0 GB	Runs comfortably	66K	67.0
Q4_K_M	4.85	4.6 GB	6.4 GB	Runs comfortably	70K	74.0
Q4_K_S	4.58	4.4 GB	6.2 GB	Runs comfortably	72K	77.3
Q4_0	4.55	4.4 GB	6.2 GB	Runs comfortably	72K	77.7
IQ4_XS	4.25	4.1 GB	5.9 GB	Runs comfortably	74K	81.7
Q3_K_M	3.91	3.8 GB	5.7 GB	Runs comfortably	77K	86.8
IQ3_M	3.70	3.7 GB	5.5 GB	Runs comfortably	78K	90.3
IQ3_XXS	3.06	3.2 GB	5.0 GB	Runs comfortably	82K	103
Q2_K	2.63	2.8 GB	4.6 GB	Runs comfortably	85K	113
IQ2_XXS	2.06	2.3 GB	4.2 GB	Runs comfortably	89K	131
IQ1_M	1.75	2.1 GB	3.9 GB	Runs comfortably	91K	144

Q4_K_M at 8K context. Bandwidth sets the speed; capacity sets the ceiling.

Model	Params	Weights	Verdict	Max ctx	tok/s
DeepSeek-R1 671B-A37B	671B	379.0 GB	Won't fit	—	1.83
Qwen3 235B-A22B	235.1B	132.8 GB	Won't fit	—	3.03
gpt-oss 120B-A5.1B	116.8B	66.0 GB	Won't fit	—	14.2
Llama 4 Scout 109B-A17B	109B	61.7 GB	Won't fit	—	4.38
Qwen2.5 72B	72.7B	41.2 GB	Won't fit	—	1.17
Llama 3.1 70B	70.5B	40.0 GB	Won't fit	—	1.21
Mixtral 8x7B	46.7B	26.4 GB	Won't fit	—	8.29
Command R 35B	35.0B	20.1 GB	Won't fit	—	1.75
Yi-1.5 34B	34.4B	19.5 GB	Won't fit	—	3.92
Qwen3 32B	32.8B	18.6 GB	Won't fit	—	4.29
Qwen2.5-Coder 32B	32.8B	18.6 GB	Won't fit	—	4.29
DeepSeek-R1-Distill-Qwen 32B	32.8B	18.6 GB	Won't fit	—	4.29
Qwen3 30B-A3B	30.5B	17.3 GB	Won't fit	—	39.6
Gemma 3 27B	27.4B	15.6 GB	Won't fit	—	7.68
Mistral Small 24B	23.6B	13.4 GB	Won't fit	—	14.3
gpt-oss 20B-A3.6B	20.9B	11.9 GB	Fits, but tight	33K	128
Qwen3 14B	14.8B	8.5 GB	Runs comfortably	32K	41.9
Phi-4 14B	14.7B	8.4 GB	Runs comfortably	16K	41.5
Mistral NeMo 12B	12.3B	7.0 GB	Runs comfortably	41K	49.6
Gemma 3 12B	12.2B	7.0 GB	Runs comfortably	98K	50.3
GLM-4 9B	9.4B	5.4 GB	Runs comfortably	128K	68.0
Qwen3 8B	8.2B	4.7 GB	Runs comfortably	62K	71.6
Llama 3.1 8B	8.0B	4.6 GB	Runs comfortably	70K	74.0
DeepSeek-R1-Distill-Llama 8B	8.0B	4.6 GB	Runs comfortably	70K	74.0
Qwen2.5 7B	7.6B	4.4 GB	Runs comfortably	128K	82.3
Mistral 7B v0.3	7.3B	4.1 GB	Runs comfortably	32K	81.9
Gemma 3 4B	4.3B	2.5 GB	Runs comfortably	128K	139
Qwen3 4B	4.0B	2.3 GB	Runs comfortably	79K	130
Llama 3.2 3B	3.2B	1.8 GB	Runs comfortably	106K	164
Llama 3.2 1B	1.2B	0.7 GB	Runs comfortably	128K	435