How much VRAM does Qwen2.5-Coder 14B need?

At Q4_K_M the weights are 8.3 GB; with an 8K context KV cache and runtime overhead it needs 10.4 GB, so a 16 GB card runs it with headroom.

Can a GeForce RTX 4070 run Qwen2.5-Coder 14B?

Yes — Q3_K_M fits with headroom at 8K context, about 45 tokens per second estimated.

Can a GeForce RTX 4090 run Qwen2.5-Coder 14B?

Yes — Q8_0 fits with headroom at 8K context, about 42 tokens per second estimated.

Can a GeForce RTX 3060 12 GB run Qwen2.5-Coder 14B?

Yes — Q3_K_M fits with headroom at 8K context, about 32 tokens per second estimated.

Can a M3 Max · 64 GB run Qwen2.5-Coder 14B?

Yes — Q8_0 fits with headroom at 8K context, about 15 tokens per second estimated.

Hardware requirements

What hardware do I need to run Qwen2.5-Coder 14B?

14.8B 128K context Apache 2.0

At Q4_K_M the weights are 8.3 GB; with an 8K KV cache and runtime overhead that is 10.4 GB, so it wants a 16 GB card to run with headroom — in practice 16 GB VRAM: GeForce RTX 5080 and 17 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 11 GB VRAM

GeForce RTX 2080 Ti — Q3_K_M, 8.8 GB of 9.7 GB, ~55 tok/s, up to 12K context.

Also: GeForce GTX 1080 Ti

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

Recommended 16 GB VRAM

GeForce RTX 5080 — Q4_K_M, 10.4 GB of 14.4 GB, ~70 tok/s, up to 29K context.

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

runs the recommended quantisation with headroom at 8K context

High quality 24 GB VRAM

GeForce RTX 4090 — Q8_0, 16.7 GB of 22.4 GB, ~42 tok/s, up to 38K context.

Also: Radeon RX 7900 XTX, GeForce RTX 3090 Ti, GeForce RTX 3090, GeForce RTX 5090 Laptop, RTX A5000 +3 more

near-lossless Q8_0 with headroom

Long context 48 GB VRAM

RTX 6000 Ada — Q4_K_M, 32.9 GB of 46.4 GB, ~70 tok/s at 128K.

Also: L40S, Radeon Pro W7900, RTX A6000

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

Apple Silicon 24 GB unified

M2 · 24 GB — Q4_K_M, 10.4 GB of 18.0 GB, ~6.7 tok/s, up to 48K context.

Also: M3 · 24 GB

smallest Mac that runs the recommended quantisation with headroom

CPU only 32 GB RAM

CPU only · DDR4 dual-channel — Q8_0, 16.7 GB of 25.6 GB, ~1.5 tok/s, up to 55K 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	27.6 GB	1.50 GB	6.00 GB	24.0 GB	29.7 GB	40 GB card	Reference
Q8_0	14.6 GB	1.50 GB	6.00 GB	24.0 GB	16.7 GB	24 GB card	−0.1% ppl
Q6_K	11.3 GB	1.50 GB	6.00 GB	24.0 GB	13.4 GB	20 GB card	−0.4% ppl
Q5_K_M	9.8 GB	1.50 GB	6.00 GB	24.0 GB	11.9 GB	16 GB card	−0.8% ppl
Q4_K_M	8.3 GB	1.50 GB	6.00 GB	24.0 GB	10.4 GB	16 GB card	−1.9% ppl
Q3_K_M	6.7 GB	1.50 GB	6.00 GB	24.0 GB	8.8 GB	12 GB card	−5.4% ppl
Q2_K	5.8 GB	1.50 GB	6.00 GB	24.0 GB	7.9 GB	11 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	114K	130	no	Check
GeForce RTX 5080	16 GB	Q5_K_M	Q6_K	29K	70	no	Check
GeForce RTX 5070 Ti	16 GB	Q5_K_M	Q6_K	29K	65	no	Check
GeForce RTX 5070	12 GB	Q3_K_M	Q4_K_M	8K	49	no	Check
GeForce RTX 5060 Ti 16 GB	16 GB	Q5_K_M	Q6_K	29K	33	no	Check
GeForce RTX 5060	8 GB	—	offload	—	~15	no	Check
GeForce RTX 4090	24 GB	Q8_0	Q8_0	71K	73	no	Check
GeForce RTX 4080 Super	16 GB	Q5_K_M	Q6_K	29K	54	no	Check
GeForce RTX 4080	16 GB	Q5_K_M	Q6_K	29K	52	no	Check
GeForce RTX 4070 Ti Super	16 GB	Q5_K_M	Q6_K	29K	49	no	Check
GeForce RTX 4070 Ti	12 GB	Q3_K_M	Q4_K_M	8K	37	no	Check
GeForce RTX 4070 Super	12 GB	Q3_K_M	Q4_K_M	8K	37	no	Check
GeForce RTX 4070	12 GB	Q3_K_M	Q4_K_M	8K	37	no	Check
GeForce RTX 4060 Ti 16 GB	16 GB	Q5_K_M	Q6_K	29K	21	no	Check
GeForce RTX 4060 Ti	8 GB	—	offload	—	~13	no	Check
GeForce RTX 4060	8 GB	—	offload	—	~12	no	Check
GeForce RTX 3090 Ti	24 GB	Q8_0	Q8_0	71K	73	no	Check
GeForce RTX 3090	24 GB	Q8_0	Q8_0	71K	68	no	Check
GeForce RTX 3080 Ti	12 GB	Q3_K_M	Q4_K_M	8K	66	no	Check
GeForce RTX 3080 12 GB	12 GB	Q3_K_M	Q4_K_M	8K	66	no	Check
GeForce RTX 3080	10 GB	—	offload	—	~64	no	Check
GeForce RTX 3070 Ti	8 GB	—	offload	—	~16	no	Check
GeForce RTX 3070	8 GB	—	offload	—	~15	no	Check
GeForce RTX 3060 Ti	8 GB	—	offload	—	~15	no	Check
GeForce RTX 3060 12 GB	12 GB	Q3_K_M	Q4_K_M	8K	26	no	Check
GeForce RTX 2080 Ti	11 GB	—	Q3_K_M	—	—	no	Check
GeForce RTX 2060 12 GB	12 GB	Q3_K_M	Q4_K_M	8K	24	no	Check
GeForce GTX 1080 Ti	11 GB	—	Q3_K_M	—	—	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	71K	65	no	Check
GeForce RTX 4090 Laptop	16 GB	Q5_K_M	Q6_K	29K	42	no	Check
GeForce RTX 4080 Laptop	12 GB	Q3_K_M	Q4_K_M	8K	31	no	Check
GeForce RTX 4070 Laptop	8 GB	—	offload	—	~12	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	70	yes	Check
RTX 5000 Ada	32 GB	Q8_0	Q8_0	114K	42	no	Check
RTX 4000 Ada	20 GB	Q6_K	Q8_0	50K	26	no	Check
RTX 2000 Ada	16 GB	Q5_K_M	Q6_K	29K	16	no	Check
RTX A6000	48 GB	Q8_0	Q8_0	128K	56	yes	Check
RTX A5000	24 GB	Q8_0	Q8_0	71K	56	no	Check
RTX A4000	16 GB	Q5_K_M	Q6_K	29K	33	no	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	63	yes	Check
L4	24 GB	Q8_0	Q8_0	71K	22	no	Check
A10	24 GB	Q8_0	Q8_0	71K	44	no	Check
A100 40 GB	40 GB	Q8_0	Q8_0	128K	113	tight	Check
A100 80 GB	80 GB	Q8_0	Q8_0	128K	148	yes	Check
H100 PCIe	80 GB	Q8_0	Q8_0	128K	145	yes	Check
H100 SXM	80 GB	Q8_0	Q8_0	128K	244	yes	Check
H200 SXM	141 GB	Q8_0	Q8_0	128K	349	yes	Check
Tesla P40	24 GB	Q8_0	Q8_0	71K	25	no	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	Q5_K_M	Q6_K	29K	47	no	Check
Radeon RX 9070	16 GB	Q5_K_M	Q6_K	29K	47	no	Check
Radeon RX 7900 XTX	24 GB	Q8_0	Q8_0	71K	70	no	Check
Radeon RX 7900 XT	20 GB	Q6_K	Q8_0	50K	58	no	Check
Radeon RX 7900 GRE	16 GB	Q5_K_M	Q6_K	29K	42	no	Check
Radeon RX 7800 XT	16 GB	Q5_K_M	Q6_K	29K	45	no	Check
Radeon RX 7700 XT	12 GB	Q3_K_M	Q4_K_M	8K	31	no	Check
Radeon RX 7600 XT	16 GB	Q5_K_M	Q6_K	29K	21	no	Check
Radeon RX 6900 XT	16 GB	Q5_K_M	Q6_K	29K	37	no	Check
Radeon RX 6800 XT	16 GB	Q5_K_M	Q6_K	29K	37	no	Check
Radeon RX 6700 XT	12 GB	Q3_K_M	Q4_K_M	8K	28	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	63	yes	Check
Radeon Pro W7800	32 GB	Q8_0	Q8_0	114K	42	no	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	119	yes	Check
Instinct MI300X	192 GB	Q8_0	Q8_0	128K	386	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	17	yes	Check
Ryzen AI Max+ 395 · 64 GB	64 GB	Q8_0	Q8_0	128K	17	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	Q3_K_M	Q4_K_M	9K	4.6	no	Check
M1 Pro · 32 GB	32 GB	Q8_0	Q8_0	80K	13	no	Check
M1 Max · 64 GB	64 GB	Q8_0	Q8_0	128K	27	yes	Check
M1 Ultra · 128 GB	128 GB	Q8_0	Q8_0	128K	54	yes	Check
M2 · 24 GB	24 GB	Q6_K	Q8_0	48K	6.7	no	Check
M2 Pro · 32 GB	32 GB	Q8_0	Q8_0	80K	13	no	Check
M2 Max · 96 GB	96 GB	Q8_0	Q8_0	128K	27	yes	Check
M2 Ultra · 192 GB	192 GB	Q8_0	Q8_0	128K	54	yes	Check
M3 · 24 GB	24 GB	Q6_K	Q8_0	48K	6.7	no	Check
M3 Pro · 36 GB	36 GB	Q8_0	Q8_0	96K	10	no	Check
M3 Max · 64 GB	64 GB	Q8_0	Q8_0	128K	27	yes	Check
M3 Max · 128 GB	128 GB	Q8_0	Q8_0	128K	27	yes	Check
M3 Ultra · 256 GB	256 GB	Q8_0	Q8_0	128K	54	yes	Check
M3 Ultra · 512 GB	512 GB	Q8_0	Q8_0	128K	54	yes	Check
M4 · 32 GB	32 GB	Q8_0	Q8_0	80K	8.1	no	Check
M4 Pro · 48 GB	48 GB	Q8_0	Q8_0	128K	18	tight	Check
M4 Pro · 64 GB	64 GB	Q8_0	Q8_0	128K	18	yes	Check
M4 Max · 64 GB	64 GB	Q8_0	Q8_0	128K	28	yes	Check
M4 Max · 128 GB	128 GB	Q8_0	Q8_0	128K	37	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	Q3_K_M	Q4_K_M	8K	33	no	Check
Arc B570	10 GB	—	offload	—	~33	no	Check
Arc A770 16 GB	16 GB	Q5_K_M	Q6_K	29K	41	no	Check
Arc A750	8 GB	—	offload	—	~15	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	88K	2.6	no	Check
CPU only · DDR5 dual-channel	64 GB	Q8_0	Q8_0	128K	4.5	yes	Check
CPU only · DDR5 8-channel server	256 GB	Q8_0	Q8_0	128K	15	yes	Check