Desktop GPU · Intel

What AI models can an Arc B580 run?

12 GB is the point where local models stop being a demo. The popular 7B to 14B models fit with room for a real conversation, and a 32B fits if you accept a lower quantization. At 456 GB/s there is enough bandwidth to keep generation responsive for anything that fits.

Among the cheapest new cards with 12 GB, and the best capacity per dollar in the budget range. It runs language models through Vulkan and SYCL builds of llama.cpp. Software support is narrower and newer than for NVIDIA, so it suits someone who is comfortable troubleshooting.

Memory12 GB

Bandwidth456 GB/s

Usable for a model11.2 GB

Runtime backendVULKAN

ArchitectureBattlemage

Power190 W

The short answer

Assuming an 8k context window and default settings, these are the models worth downloading first.

Best all-rounder

Gemma 4 12B Instruct

Q6_K · 10.84 GB · about 24.8 tokens/s

Best for code

Qwen2.5 Coder 14B Instruct

Q4_K_M · 9.81 GB · about 27.7 tokens/s

Largest that still runs well

Qwen3 14B

14.77B parameters · Q5_K_M · about 24.6 tokens/s

Fastest useful answer

LFM2.5 8B A1B

about 91.2 tokens/s · 8.81 GB

See how fast it feels

Arc B580 running Gemma 4 12B Instruct at Q6_K

Wait for the first word1.3 s

Then writes at24.8 tok/s

Whole answer8.3 s

YouWhy does my model use more memory when the conversation gets longer?

Model

Simulated from our estimate at a 512-token question, not a recording. Assumes nothing else is competing for the GPU.

Where it sits

With 12 GB it belongs to the 12 GB tier. The 12 GB guide explains what that tier runs and where it runs out, and the bandwidth guide explains why 456 GB/s decides how fast it answers.

Our indicative price for this card is $260 new, reviewed 2026-09-01. That is 4.6 GB of memory and 175 GB/s of bandwidth for every $100. Prices move, so treat these as a place to start a search and not a quote.

Devices in a similar price band, fastest memory first:

Device	Price	Memory	Bandwidth	Best all-round pick
GeForce RTX 2080 Ti	$250 used	11 GB	616 GB/s	Qwen3.5 9B, 51.3 t/s
GeForce RTX 3070 Ti	$300 used	8 GB	608 GB/s	Qwen3.5 9B, 74.4 t/s
Arc A770 16GB	$280 used	16 GB	560 GB/s	Qwen3.5 27B, 23.1 t/s
Arc A750	$200 used	8 GB	512 GB/s	Qwen3.5 9B, 48.3 t/s

Can an Arc B580 run it?

A direct answer for the models people ask about most, each with the settings we would use.

Qwen2.5 7B Instruct Llama 3.1 8B Instruct Qwen3 8B Qwen3.5 9B Gemma 4 12B Instruct Qwen3 14B Phi 4 GPT OSS 20B Qwen3.5 27B Gemma 4 31B Instruct Qwen3 32B Qwen3 30B A3B GPT OSS 120B

Every model, scored on this device

Each row uses the highest-quality quantization that both fits and stays conversational. Speed is a single-stream estimate at 8k context.

Model	Params	Verdict	Download	Memory used	Speed	Max context
Gemma 4 12B Instruct Gemma	11.96B	Just fits	Q6_K	10.84 GB	24.8 t/s	8k
Qwen3.5 9B Qwen	9.65B	Runs great	Q8_0	9.25 GB	29.3 t/s	64k
Qwen3 14B Qwen	14.77B	Just fits	Q5_K_M	10.96 GB	24.6 t/s	8k
Gemma 3 12B Instruct Gemma	12.19B	Just fits	Q6_K	10.52 GB	25.6 t/s	16k
Gemma 4 E4B Instruct Gemma	8B	Runs great	Q8_0	8.42 GB	32.1 t/s	128k
Granite 4.1 8B Granite	8.79B	Just fits	Q8_0	10.35 GB	26 t/s	8k
Qwen2.5 14B Instruct Qwen	14.77B	Runs great	Q4_K_M	9.81 GB	27.7 t/s	8k
Qwen3 8B Qwen	8.19B	Runs great	Q8_0	9.02 GB	30.1 t/s	16k
Qwen2.5 Coder 14B Instruct Qwen	14.77B	Runs great	Q4_K_M	9.81 GB	27.7 t/s	8k
Qwen2.5 VL 7B Instruct Qwen	8.29B	Runs great	Q8_0	8.49 GB	32 t/s	32k
Phi 4 Phi	14.66B	Runs great	Q4_K_M	9.96 GB	27.2 t/s	8k
Mistral Nemo Instruct 2407 Mistral	12.25B	Just fits	Q6_K	10.53 GB	25.7 t/s	8k
Qwen3.5 27B Qwen	27.78B	Fits only at 2 to 3 bits	IQ3_XXS !	10.53 GB	25.7 t/s	16k
Qwen2.5 7B Instruct Qwen	7.62B	Runs great	Q8_0	7.83 GB	34.8 t/s	32k
Llama 3.1 8B Instruct Llama	8.03B	Runs great	Q8_0	8.83 GB	30.7 t/s	16k
Qwen2.5 Coder 7B Instruct Qwen	7.62B	Runs great	Q8_0	7.83 GB	34.8 t/s	32k
Gemma 2 9B Instruct Gemma	9.24B	Runs great	Q6_K	9.42 GB	28.7 t/s	8k
OLMo 3 7B Instruct OLMo	7.3B	Runs great	Q8_0	9.72 GB	27.8 t/s	16k
Internlm3 8B Instruct InternLM	8.8B	Runs great	Q8_0	8.97 GB	30.3 t/s	32k
Mistral 7B Instruct V0.3 Mistral	7.25B	Runs great	Q8_0	8.45 GB	32.2 t/s	16k
DeepSeek Coder V2 Lite Instruct DeepSeek	15.71B (2.74B active)	Just fits	Q5_K_M	10.84 GB	74.1 t/s	16k
Gemma 4 E2B Instruct Gemma	5.12B	Runs great	Q8_0	5.44 GB	50.6 t/s	128k
Qwen3.5 4B Qwen	4.66B	Runs great	Q8_0	5.27 GB	52.8 t/s	128k
Mistral 7B Instruct V0.2 Mistral	7.24B	Runs great	Q8_0	8.44 GB	32.2 t/s	16k
DeepSeek Coder 7B Instruct V1.5 DeepSeek	6.91B	Just fits	Q8_0	10.58 GB	25.5 t/s	4k
Mistral Small 24B Instruct 2501 Mistral	23.57B	Fits only at 2 to 3 bits	IQ3_XXS !	9.85 GB	27.5 t/s	16k
DeepSeek Coder 6.7B Instruct DeepSeek	6.74B	Just fits	Q8_0	10.94 GB	24.6 t/s	8k
LFM2.5 8B A1B Liquid	8.47B (1.57B active)	Runs great	Q8_0	8.81 GB	91.2 t/s	64k
Qwen3 4B Qwen	4.02B	Runs great	Q8_0	5.46 GB	50.8 t/s	32k
Codestral 22B V0.1 Mistral	22.25B	Fits only at 2 to 3 bits	IQ3_XXS !	10.08 GB	27 t/s	8k

Show the other 56 models

Gemma 3 4B Instruct	4.3B	Runs great	Q8_0	4.91 GB	56.9 t/s	128k
Phi 4 Mini Instruct	3.84B	Runs great	Q8_0	5.17 GB	54 t/s	32k
Granite 4.1 3B	3.4B	Runs great	Q8_0	4.34 GB	64.9 t/s	64k
Qwen3 32B	32.76B	Fits only at 2 to 3 bits	IQ2_XXS	10.11 GB	26.8 t/s	8k
Qwen2.5 3B Instruct	3.09B	Runs great	Q8_0	3.68 GB	77.6 t/s	32k
Phi 3 Mini 4k Instruct	3.82B	Runs great	Q8_0	4.8 GB	58.4 t/s	4k
Llama 3.2 3B Instruct	3.21B	Runs great	Q8_0	4.42 GB	63.9 t/s	64k
Granite 4.1 30B	28.87B	Fits only at 2 to 3 bits	IQ2_XXS	9.33 GB	29 t/s	8k
Qwen2.5 32B Instruct	32.76B	Fits only at 2 to 3 bits	IQ2_XXS	10.11 GB	26.8 t/s	8k
Qwen2.5 Coder 32B Instruct	32.76B	Fits only at 2 to 3 bits	IQ2_XXS	10.11 GB	26.8 t/s	8k
LFM2.5 2.6B	2.7B	Runs great	Q8_0	3.14 GB	92.6 t/s	128k
Qwen3.5 2B	2.27B	Runs great	Q8_0	2.7 GB	109.9 t/s	256k
Gemma 3 27B Instruct	27.43B	Fits only at 2 to 3 bits	IQ2_XXS	8.11 GB	33.8 t/s	32k
Gemma 2 2B Instruct	2.61B	Runs great	Q8_0	3.54 GB	81.2 t/s	8k
Starcoder2 3B	3.03B	Runs great	Q8_0	3.34 GB	87.3 t/s	16k
Qwen3 1.7B	2.03B	Runs great	Q8_0	3.22 GB	89.8 t/s	32k
DeepSeek R1 Distill Qwen 1.5B	1.78B	Runs great	Q8_0	2.07 GB	148.1 t/s	128k
Qwen2.5 1.5B Instruct	1.54B	Runs great	Q8_0	2.06 GB	148.7 t/s	32k
SmolLM2 1.7B	1.71B	Runs great	Q8_0	3.53 GB	81.2 t/s	8k
OLMo 2 0425 1B	1.48B	Runs great	Q8_0	2.6 GB	114.6 t/s	4k
Llama 3.2 1B Instruct	1.24B	Runs great	Q8_0	1.82 GB	175.4 t/s	128k
LFM2.5 1.2B Instruct	1.17B	Runs great	Q8_0	1.46 GB	231.6 t/s	64k
Gemma 3 1B Instruct	1B	Runs great	Q8_0	1.35 GB	249.4 t/s	32k
Qwen3.5 0.8B	0.87B	Runs great	Q8_0	1.28 GB	266.3 t/s	256k
Qwen3 0.6B	0.75B	Runs great	Q8_0	1.92 GB	160.2 t/s	32k
Qwen2.5 0.5B Instruct	0.49B	Runs great	Q8_0	0.88 GB	447.7 t/s	32k
LFM2.5 350M	0.35B	Runs great	Q8_0	0.68 GB	693.3 t/s	64k
GPT OSS 20B	20.91B	Partial offload	n/a	11.55 GB	64.3 t/s	n/a
Qwen3.5 35B A3B	35.95B	Partial offload	n/a	20.48 GB	55.1 t/s	n/a
GLM 4.7 Flash	31.22B	Partial offload	n/a	18.13 GB	37.2 t/s	n/a
Gemma 4 26B A4B Instruct	25.81B	Partial offload	n/a	15.7 GB	37.1 t/s	n/a
Qwen3 30B A3B	30.53B	Partial offload	n/a	18.08 GB	35.9 t/s	n/a
Qwen3 Coder 30B A3B Instruct	30.53B	Partial offload	n/a	18.08 GB	35.9 t/s	n/a
Qwen3 Coder Next	79.67B	Partial offload	n/a	45.22 GB	33 t/s	n/a
Qwen3 Next 80B A3B Instruct	81.32B	Partial offload	n/a	46.15 GB	32.5 t/s	n/a
GPT OSS 120B	116.83B	Partial offload	n/a	59.4 GB	16.8 t/s	n/a
Phi 3.5 MoE Instruct	41.87B	Partial offload	n/a	24.87 GB	13 t/s	n/a
Qwen3.5 122B A10B	125.09B	Partial offload	n/a	70.61 GB	12.6 t/s	n/a
GLM 4.5 Air	110.47B	Partial offload	n/a	63.61 GB	6.9 t/s	n/a
MiniMax M2.5	228.7B	Partial offload	n/a	130.56 GB	6.6 t/s	n/a
MiniMax M2.7	228.69B	Partial offload	n/a	130.55 GB	6.6 t/s	n/a
Mixtral 8x7B Instruct V0.1	46.7B	Partial offload	n/a	27.59 GB	6.1 t/s	n/a
MiniMax M3	427.04B	Partial offload	n/a	240.84 GB	3.4 t/s	n/a
Qwen3 235B A22B	235.09B	Partial offload	n/a	133.71 GB	3.4 t/s	n/a
Gemma 4 31B Instruct	31.27B	Partial offload	n/a	21.31 GB	2.6 t/s	n/a
Kimi K2.6	1026.88B	Partial offload	n/a	577.78 GB	2.2 t/s	n/a
Kimi K2 Instruct	1026.41B	Partial offload	n/a	577.52 GB	2.2 t/s	n/a
DeepSeek V3.2	685.4B	Partial offload	n/a	385.91 GB	2.2 t/s	n/a
DeepSeek R1	684.53B	Partial offload	n/a	385.42 GB	2.2 t/s	n/a
GLM 5.1	753.86B	Partial offload	n/a	424.51 GB	2.1 t/s	n/a
GLM 4.5	358.34B	Partial offload	n/a	204.37 GB	1.4 t/s	n/a
GLM 5.2	753.33B	Partial offload	n/a	424.21 GB	1 t/s	n/a
Qwen2.5 VL 72B Instruct	73.41B	Partial offload	n/a	43.62 GB	0.9 t/s	n/a
Qwen2.5 72B Instruct	72.71B	Partial offload	n/a	43.22 GB	0.9 t/s	n/a
Llama 3.3 70B Instruct	70.55B	Partial offload	n/a	42.12 GB	0.9 t/s	n/a
Llama 3 3 Nemotron Super 49B V1	49.87B	Partial offload	n/a	47.99 GB	0.8 t/s	n/a

What it will not run

0 of the 86 architectures we track are out of reach here, even at 2-bit precision. Another 29 run by splitting the model between the card and system RAM, which works but drops generation to single digits for most dense models.

If you want more headroom

The next steps up in memory, in order. More memory changes which models load at all; more bandwidth changes how fast they answer, so the two are worth weighing separately.

Device	Memory	Bandwidth	Models that fit
GeForce RTX 5080	16 GB	960 GB/s	63	Check price
GeForce RTX 5070 Ti	16 GB	896 GB/s	63	Check price
GeForce RTX 4080 SUPER	16 GB	736 GB/s	63	Check price
GeForce RTX 4080	16 GB	717 GB/s	63	Check price

Price links go to an Amazon search for the model name and are affiliate links: if you buy through one, we earn a commission at no cost to you. We do not take payment for placement, and the ordering above is by memory capacity alone.