How much VRAM does Qwen2.5-Coder 7B need?

At Q4_K_M the weights are 4.3 GB; with an 8K context KV cache and runtime overhead it needs 5.3 GB, so a 8 GB card runs it with headroom.

Can a GeForce RTX 4070 run Qwen2.5-Coder 7B?

Yes — Q8_0 fits with headroom at 8K context, about 40 tokens per second estimated.

Can a GeForce RTX 4090 run Qwen2.5-Coder 7B?

Yes — Q8_0 fits with headroom at 8K context, about 81 tokens per second estimated.

Can a GeForce RTX 3060 12 GB run Qwen2.5-Coder 7B?

Yes — Q8_0 fits with headroom at 8K context, about 29 tokens per second estimated.

Can a M3 Max · 64 GB run Qwen2.5-Coder 7B?

Yes — Q8_0 fits with headroom at 8K context, about 30 tokens per second estimated.

Hardware requirements

What hardware do I need to run Qwen2.5-Coder 7B?

7.62B 128K context Apache 2.0

At Q4_K_M the weights are 4.3 GB; with an 8K KV cache and runtime overhead that is 5.3 GB, so it wants a 8 GB card to run with headroom — in practice 8 GB VRAM: GeForce RTX 3070 Ti and 7 others. Every figure below is computed from the model’s published shape against each device’s usable memory; tokens per second are estimates. The method.

Minimum practical 8 GB VRAM

GeForce RTX 3070 Ti — Q6_K, 6.9 GB of 7.0 GB, ~63 tok/s, up to 10K context.

Also: Arc A750, GeForce RTX 5060, GeForce RTX 3070, GeForce RTX 3060 Ti, GeForce RTX 4060 Ti +2 more

the smallest card that holds it at all, at Q3_K_M — the lowest quantisation worth running

Recommended 8 GB VRAM

GeForce RTX 3070 Ti — Q4_K_M, 5.3 GB of 7.0 GB, ~86 tok/s, up to 38K context.

Also: Arc A750, GeForce RTX 5060, GeForce RTX 3070, GeForce RTX 3060 Ti, GeForce RTX 4060 Ti +2 more

runs the recommended quantisation with headroom at 8K context

High quality 12 GB VRAM

GeForce RTX 3060 12 GB — Q8_0, 8.6 GB of 10.6 GB, ~29 tok/s, up to 44K context.

Also: GeForce RTX 3080 Ti, GeForce RTX 3080 12 GB, GeForce RTX 5070, GeForce RTX 4070 Ti, GeForce RTX 4070 Super +6 more

near-lossless Q8_0 with headroom

Long context 16 GB VRAM

GeForce RTX 5080 — Q4_K_M, 11.9 GB of 14.4 GB, ~136 tok/s at 128K.

Also: GeForce RTX 5070 Ti, GeForce RTX 4080 Super, GeForce RTX 4080, GeForce RTX 4070 Ti Super, Radeon RX 9070 XT +12 more

128K tokens (or the model’s maximum) at the recommended quantisation

Apple Silicon 16 GB unified

M1 · 16 GB — Q4_K_M, 5.3 GB of 10.7 GB, ~8.9 tok/s, up to 106K context.

smallest Mac that runs the recommended quantisation with headroom

CPU only 32 GB RAM

CPU only · DDR4 dual-channel — Q8_0, 8.6 GB of 25.6 GB, ~2.8 tok/s, up to 128K context.

no GPU — weights stream from system RAM

How much VRAM at each quantisation

Quant	Weights	+KV 8K	+KV 32K	+KV 128K	Total @ 8K	Needs	Quality
F16	14.2 GB	0.44 GB	1.75 GB	7.0 GB	15.2 GB	20 GB card	Reference
Q8_0	7.5 GB	0.44 GB	1.75 GB	7.0 GB	8.6 GB	12 GB card	−0.1% ppl
Q6_K	5.8 GB	0.44 GB	1.75 GB	7.0 GB	6.9 GB	10 GB card	−0.4% ppl
Q5_K_M	5.0 GB	0.44 GB	1.75 GB	7.0 GB	6.1 GB	10 GB card	−0.8% ppl
Q4_K_M	4.3 GB	0.44 GB	1.75 GB	7.0 GB	5.3 GB	8 GB card	−1.9% ppl
Q3_K_M	3.5 GB	0.44 GB	1.75 GB	7.0 GB	4.5 GB	8 GB card	−5.4% ppl
Q2_K	3.0 GB	0.44 GB	1.75 GB	7.0 GB	4.0 GB	8 GB card	−15% ppl

Totals add 0.6 GB runtime overhead and an f16 KV cache; "needs" is the smallest discrete card class whose usable memory (after display and driver reserve) leaves 15% headroom. A q8_0 cache roughly halves the KV columns.

Every device, checked

System RAM for offload:

16 32 64 128 256

NVIDIA · GeForce

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
GeForce RTX 5090	32 GB	Q8_0	Q8_0	128K	253	yes	Check
GeForce RTX 5080	16 GB	Q8_0	Q8_0	128K	136	yes	Check
GeForce RTX 5070 Ti	16 GB	Q8_0	Q8_0	128K	127	yes	Check
GeForce RTX 5070	12 GB	Q8_0	Q8_0	104K	95	no	Check
GeForce RTX 5060 Ti 16 GB	16 GB	Q8_0	Q8_0	128K	63	yes	Check
GeForce RTX 5060	8 GB	Q4_K_M	Q6_K	38K	63	no	Check
GeForce RTX 4090	24 GB	Q8_0	Q8_0	128K	142	yes	Check
GeForce RTX 4080 Super	16 GB	Q8_0	Q8_0	128K	104	yes	Check
GeForce RTX 4080	16 GB	Q8_0	Q8_0	128K	101	yes	Check
GeForce RTX 4070 Ti Super	16 GB	Q8_0	Q8_0	128K	95	yes	Check
GeForce RTX 4070 Ti	12 GB	Q8_0	Q8_0	104K	71	no	Check
GeForce RTX 4070 Super	12 GB	Q8_0	Q8_0	104K	71	no	Check
GeForce RTX 4070	12 GB	Q8_0	Q8_0	104K	71	no	Check
GeForce RTX 4060 Ti 16 GB	16 GB	Q8_0	Q8_0	128K	41	yes	Check
GeForce RTX 4060 Ti	8 GB	Q4_K_M	Q6_K	38K	41	no	Check
GeForce RTX 4060	8 GB	Q4_K_M	Q6_K	38K	38	no	Check
GeForce RTX 3090 Ti	24 GB	Q8_0	Q8_0	128K	142	yes	Check
GeForce RTX 3090	24 GB	Q8_0	Q8_0	128K	132	yes	Check
GeForce RTX 3080 Ti	12 GB	Q8_0	Q8_0	104K	129	no	Check
GeForce RTX 3080 12 GB	12 GB	Q8_0	Q8_0	104K	129	no	Check
GeForce RTX 3080	10 GB	Q6_K	Q8_0	71K	107	no	Check
GeForce RTX 3070 Ti	8 GB	Q4_K_M	Q6_K	38K	86	no	Check
GeForce RTX 3070	8 GB	Q4_K_M	Q6_K	38K	63	no	Check
GeForce RTX 3060 Ti	8 GB	Q4_K_M	Q6_K	38K	63	no	Check
GeForce RTX 3060 12 GB	12 GB	Q8_0	Q8_0	104K	51	no	Check
GeForce RTX 2080 Ti	11 GB	Q6_K	Q8_0	88K	87	no	Check
GeForce RTX 2060 12 GB	12 GB	Q8_0	Q8_0	104K	47	no	Check
GeForce GTX 1080 Ti	11 GB	Q6_K	Q8_0	88K	68	no	Check

NVIDIA · GeForce laptop

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
GeForce RTX 5090 Laptop	24 GB	Q8_0	Q8_0	128K	127	yes	Check
GeForce RTX 4090 Laptop	16 GB	Q8_0	Q8_0	128K	81	yes	Check
GeForce RTX 4080 Laptop	12 GB	Q8_0	Q8_0	104K	61	no	Check
GeForce RTX 4070 Laptop	8 GB	Q4_K_M	Q6_K	38K	36	no	Check

NVIDIA · Workstation

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
RTX 6000 Ada	48 GB	Q8_0	Q8_0	128K	136	yes	Check
RTX 5000 Ada	32 GB	Q8_0	Q8_0	128K	81	yes	Check
RTX 4000 Ada	20 GB	Q8_0	Q8_0	128K	51	yes	Check
RTX 2000 Ada	16 GB	Q8_0	Q8_0	128K	32	yes	Check
RTX A6000	48 GB	Q8_0	Q8_0	128K	109	yes	Check
RTX A5000	24 GB	Q8_0	Q8_0	128K	109	yes	Check
RTX A4000	16 GB	Q8_0	Q8_0	128K	63	yes	Check

NVIDIA · Data centre

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
L40S	48 GB	Q8_0	Q8_0	128K	122	yes	Check
L4	24 GB	Q8_0	Q8_0	128K	42	yes	Check
A10	24 GB	Q8_0	Q8_0	128K	85	yes	Check
A100 40 GB	40 GB	Q8_0	Q8_0	128K	220	yes	Check
A100 80 GB	80 GB	Q8_0	Q8_0	128K	288	yes	Check
H100 PCIe	80 GB	Q8_0	Q8_0	128K	283	yes	Check
H100 SXM	80 GB	Q8_0	Q8_0	128K	473	yes	Check
H200 SXM	141 GB	Q8_0	Q8_0	128K	678	yes	Check
Tesla P40	24 GB	Q8_0	Q8_0	128K	49	yes	Check

AMD · Radeon

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
Radeon RX 9070 XT	16 GB	Q8_0	Q8_0	128K	91	yes	Check
Radeon RX 9070	16 GB	Q8_0	Q8_0	128K	91	yes	Check
Radeon RX 7900 XTX	24 GB	Q8_0	Q8_0	128K	136	yes	Check
Radeon RX 7900 XT	20 GB	Q8_0	Q8_0	128K	113	yes	Check
Radeon RX 7900 GRE	16 GB	Q8_0	Q8_0	128K	81	yes	Check
Radeon RX 7800 XT	16 GB	Q8_0	Q8_0	128K	88	yes	Check
Radeon RX 7700 XT	12 GB	Q8_0	Q8_0	104K	61	no	Check
Radeon RX 7600 XT	16 GB	Q8_0	Q8_0	128K	41	yes	Check
Radeon RX 6900 XT	16 GB	Q8_0	Q8_0	128K	72	yes	Check
Radeon RX 6800 XT	16 GB	Q8_0	Q8_0	128K	72	yes	Check
Radeon RX 6700 XT	12 GB	Q8_0	Q8_0	104K	54	no	Check

AMD · Radeon Pro

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
Radeon Pro W7900	48 GB	Q8_0	Q8_0	128K	122	yes	Check
Radeon Pro W7800	32 GB	Q8_0	Q8_0	128K	81	yes	Check

AMD · Instinct

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
Instinct MI210	64 GB	Q8_0	Q8_0	128K	231	yes	Check
Instinct MI300X	192 GB	Q8_0	Q8_0	128K	749	yes	Check

AMD · Ryzen AI Max

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
Ryzen AI Max+ 395 · 128 GB	128 GB	Q8_0	Q8_0	128K	33	yes	Check
Ryzen AI Max+ 395 · 64 GB	64 GB	Q8_0	Q8_0	128K	33	yes	Check

Apple · Apple Silicon

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
M1 · 16 GB	16 GB	Q8_0	Q8_0	106K	8.9	no	Check
M1 Pro · 32 GB	32 GB	Q8_0	Q8_0	128K	26	yes	Check
M1 Max · 64 GB	64 GB	Q8_0	Q8_0	128K	52	yes	Check
M1 Ultra · 128 GB	128 GB	Q8_0	Q8_0	128K	104	yes	Check
M2 · 24 GB	24 GB	Q8_0	Q8_0	128K	13	yes	Check
M2 Pro · 32 GB	32 GB	Q8_0	Q8_0	128K	26	yes	Check
M2 Max · 96 GB	96 GB	Q8_0	Q8_0	128K	52	yes	Check
M2 Ultra · 192 GB	192 GB	Q8_0	Q8_0	128K	104	yes	Check
M3 · 24 GB	24 GB	Q8_0	Q8_0	128K	13	yes	Check
M3 Pro · 36 GB	36 GB	Q8_0	Q8_0	128K	20	yes	Check
M3 Max · 64 GB	64 GB	Q8_0	Q8_0	128K	52	yes	Check
M3 Max · 128 GB	128 GB	Q8_0	Q8_0	128K	52	yes	Check
M3 Ultra · 256 GB	256 GB	Q8_0	Q8_0	128K	104	yes	Check
M3 Ultra · 512 GB	512 GB	Q8_0	Q8_0	128K	104	yes	Check
M4 · 32 GB	32 GB	Q8_0	Q8_0	128K	16	yes	Check
M4 Pro · 48 GB	48 GB	Q8_0	Q8_0	128K	36	yes	Check
M4 Pro · 64 GB	64 GB	Q8_0	Q8_0	128K	36	yes	Check
M4 Max · 64 GB	64 GB	Q8_0	Q8_0	128K	53	yes	Check
M4 Max · 128 GB	128 GB	Q8_0	Q8_0	128K	71	yes	Check

Intel · Arc

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
Arc B580	12 GB	Q8_0	Q8_0	104K	64	no	Check
Arc B570	10 GB	Q6_K	Q8_0	71K	54	no	Check
Arc A770 16 GB	16 GB	Q8_0	Q8_0	128K	79	yes	Check
Arc A750	8 GB	Q4_K_M	Q6_K	38K	72	no	Check

CPU · CPU + system RAM

Device	Memory	Best quant with headroom	Fits at all	Max context @ Q4_K_M	Tok/s est.	128K
CPU only · DDR4 dual-channel	32 GB	Q8_0	Q8_0	128K	5.0	yes	Check
CPU only · DDR5 dual-channel	64 GB	Q8_0	Q8_0	128K	8.8	yes	Check
CPU only · DDR5 8-channel server	256 GB	Q8_0	Q8_0	128K	30	yes	Check