Best Local AI Models for Intel Arc B580 (12 GB VRAM)

Q: What is the VRAM and memory bandwidth of the Intel Arc B580?

The Intel Arc B580 features 12 GB of VRAM and a memory bandwidth of 456 GB/s (GDDR6 192-bit). In local language model inference, VRAM determines which model sizes fit on the card, while memory bandwidth dictates how many tokens per second the GPU generates.

Q: What is the best local AI model to run on a Intel Arc B580?

The best overall model for the Intel Arc B580 is Qwen/Qwen3-8B-GGUF using the recommended Q8_0 quantization (9.47 GB total memory). With 12 GB of VRAM, this GPU typically runs a 14B model at Q5, comfortably at full GPU speed. Check our guide to the best LLMs for 12 GB VRAM for details.

The Intel Arc B580 has 12 GB of VRAM. Here are the popular AI models it can run locally (4,096-token context, ~16.0 GB system RAM assumed), ranked by popularity.

Verdict: Next-gen Intel Battlemage: 12 GB VRAM with strong compute for affordable local AI.

VRAM & Memory

12 GB

GDDR6 192-bit

Bandwidth

456 GB/s

Battlemage

Fits in VRAM

33 models

Zero offload

TDP / Assumed RAM

190W

RAM: 16.0 GB

At a glance — 33 of the 36 most popular models run fully in VRAM on the Intel Arc B580 (12 GB), with about 76 tokens/s estimated on an 8B model in Q4. A solid pick is Qwen/Qwen3-8B-GGUF (Q8_0, 9.47 GB, ~56.2 t/s estimated).

The Intel Arc B580 is built on Intel's Battlemage architecture featuring GDDR6 192-bit delivering 456 GB/s of raw memory bandwidth. Next-gen Intel Battlemage: 12 GB VRAM with strong compute for affordable local AI.

Equipped with 12 GB of dedicated VRAM, the Intel Arc B580 can run 33 popular open-source models completely in GPU memory without offloading. This includes full-speed execution for weights like Qwen3-8B-GGUF, Qwen3-4B-GGUF, Qwen2.5-7B-Instruct-GGUF. For a comprehensive breakdown of compatible model weights, see our guide to the best LLMs for 12 GB VRAM.

With a memory bandwidth of 456 GB/s, this card can generate tokens at an estimated peak rate of ~76.0 tokens/second on an 8B parameter model (Q4_K_M). Its rated power draw is 190W TDP, so ensure your system's power supply and case ventilation are adequate for sustained local inferencing.

New to this? Read: How much VRAM do you need?

How the Intel Arc B580 compares with similar cards

Compared with the AMD Radeon RX 6750 XT, the Intel Arc B580 has the same 12 GB of VRAM, so the same models fit fully on the GPU; about 6% faster token generation (~76 vs ~72 tokens/s on an 8B model in Q4, from 456 vs 432 GB/s); a lower board power (190 vs 250 W) and a different architecture (Battlemage vs RDNA 2). Side-by-side comparison
Compared with the AMD Radeon RX 7700 XT, the Intel Arc B580 has the same 12 GB of VRAM, so the same models fit fully on the GPU; about 6% faster token generation (~76 vs ~72 tokens/s on an 8B model in Q4, from 456 vs 432 GB/s); a lower board power (190 vs 245 W) and a different architecture (Battlemage vs RDNA 3). Side-by-side comparison
Compared with the NVIDIA RTX 4070, the Intel Arc B580 has the same 12 GB of VRAM, so the same models fit fully on the GPU; about 10% slower token generation (~76 vs ~84 tokens/s on an 8B model in Q4, from 456 vs 504 GB/s); a lower board power (190 vs 200 W) and a different architecture (Battlemage vs Ada Lovelace). Side-by-side comparison

Token speeds here are a conservative estimate (memory bandwidth ÷ 6, for an 8B model in Q4 that reads about 5 GB of weights per token, plus overhead); prompt processing depends on GPU compute instead.

33 fit fully in VRAM · 3 run with offload

Only show models that fit fully in VRAM

Model	Size	Quant.	Quality	Memory	Speed~	Verdict
Qwen/Qwen3-8B-GGUF	8.19B	Q8_0	Excellent	9.47 GB	56.2 t/s	Fits in VRAM
Qwen/Qwen3-4B-GGUF	4.02B	Q8_0	Excellent	5.35 GB	114.4 t/s	Fits in VRAM
Qwen/Qwen2.5-7B-Instruct-GGUF	7.62B	Q8_0	Excellent	8.56 GB	60.5 t/s	Fits in VRAM
unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF	8.0B	Q8_K_XL	Excellent	8.48 GB	59.8 t/s	Fits in VRAM
unsloth/gpt-oss-20b-GGUF	20.91B	Q4_K_XL	Good	11.95 GB	41.2 t/s	Fits in VRAM
bartowski/Qwen2.5-32B-Instruct-GGUF	32.76B	IQ2_S	Very low	11.47 GB	47.1 t/s	Fits in VRAM
Qwen/Qwen2.5-1.5B-Instruct-GGUF	1.54B	GGUF	Excellent	4.23 GB	137.5 t/s	Fits in VRAM
unsloth/LFM2.5-24B-A2B-GGUF	23.97B	Q3_K_XL	Fair	11.07 GB	44.7 t/s	Fits in VRAM
unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF	30.53B	Q2_K_L	Low	11.73 GB	43.2 t/s	Fits in VRAM
Qwen/Qwen2.5-0.5B-Instruct-GGUF	0.49B	GGUF	Excellent	2.03 GB	386.6 t/s	Fits in VRAM
unsloth/Ornith-1.0-9B-GGUF	—	Q8_0	Excellent	9.8 GB	51.4 t/s	Fits in VRAM
Qwen/Qwen3-14B-GGUF	14.77B	Q5_K_M	Very good	11.22 GB	46.6 t/s	Fits in VRAM
Qwen/Qwen2.5-Coder-7B-Instruct-GGUF	7.62B	Q8_0	Excellent	8.56 GB	60.5 t/s	Fits in VRAM
ggml-org/gpt-oss-120b-GGUF	116.83B	BF16	Very low	2.42 GB	308.6 t/s	Fits in VRAM
LiquidAI/LFM3-1.7B-GGUF	1.73B	BF16	Excellent	4.08 GB	141.2 t/s	Fits in VRAM
unsloth/flint-GGUF	39.79B	IQ2_XXS	Very low	11.95 GB	41.3 t/s	Fits in VRAM
janhq/Jan-v3.5-4B-gguf	4.41B	GGUF	Excellent	9.59 GB	55.5 t/s	Fits in VRAM
Qwen/Qwen2.5-3B-Instruct-GGUF	3.09B	GGUF	Excellent	7.27 GB	72.0 t/s	Fits in VRAM
unsloth/GLM-4.7-Flash-GGUF	31.22B	Q2_K_L	Low	11.64 GB	42.9 t/s	Fits in VRAM
bartowski/Meta-Llama-3.1-8B-Instruct-GGUF	8.03B	Q8_0	Excellent	9.23 GB	57.3 t/s	Fits in VRAM
unsloth/Llama-3.2-3B-Instruct-GGUF	3.21B	F16	Excellent	7.23 GB	76.1 t/s	Fits in VRAM
LiquidAI/LFM2.5-1.2B-Instruct-GGUF	1.17B	BF16	Excellent	3.03 GB	208.9 t/s	Fits in VRAM
lmstudio-community/gemma-3-1b-it-GGUF	1.0B	Q8_0	Excellent	1.96 GB	457.9 t/s	Fits in VRAM
bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF	34.66B	IQ2_S	Very low	11.13 GB	44.5 t/s	Fits in VRAM
Qwen/Qwen3-0.6B-GGUF	0.75B	Q8_0	Excellent	1.83 GB	765.7 t/s	Fits in VRAM
bartowski/Qwen_Qwen3-0.6B-GGUF	0.75B	BF16	Excellent	2.64 GB	324.4 t/s	Fits in VRAM
unsloth/Qwen3-1.7B-GGUF	1.72B	BF16	Excellent	4.45 GB	142.0 t/s	Fits in VRAM
LiquidAI/LFM2.5-350M-GGUF	0.35B	BF16	Excellent	1.51 GB	688.2 t/s	Fits in VRAM
Qwen/Qwen2.5-Coder-14B-Instruct-GGUF	14.77B	Q5_K_M	Very good	11.34 GB	46.6 t/s	Fits in VRAM
unsloth/DeepSeek-R1-0528-Qwen3-8B-GGUF	8.19B	Q8_K_XL	Excellent	11.44 GB	45.2 t/s	Fits in VRAM
bartowski/Nanbeige_Nanbeige4.3-33B-GGUF	33.12B	IQ2_S	Low	12.0 GB	43.9 t/s	Fits in VRAM
Qwen/Qwen2.5-Coder-1.5B-Instruct-GGUF	1.54B	Q8_0	Excellent	2.67 GB	258.4 t/s	Fits in VRAM
bartowski/Qwen2.5-14B-Instruct-GGUF	14.77B	Q5_K_L	Very good	11.78 GB	44.6 t/s	Fits in VRAM
ggml-org/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF	31.58B	Q4_0	Good	18.42 GB	2.8 t/s	Offload
bartowski/Llama-3.3-70B-Instruct-GGUF	70.55B	Q2_K_L	Low	27.74 GB	2.0 t/s	Offload
unsloth/Qwen3-Coder-Next-GGUF	79.67B	IQ3_XXS	Low	27.7 GB	1.9 t/s	Offload

"Fits in VRAM" = fast, fully on GPU. "Offload" = part on system RAM, slower. Speed is a rough estimate from the Intel Arc B580's 456 GB/s memory bandwidth (offloaded models are limited by system RAM).

Frequently asked questions

What is the VRAM and memory bandwidth of the Intel Arc B580?

The Intel Arc B580 features 12 GB of VRAM and a memory bandwidth of 456 GB/s (GDDR6 192-bit). In local language model inference, VRAM determines which model sizes fit on the card, while memory bandwidth dictates how many tokens per second the GPU generates.

What is the best local AI model to run on a Intel Arc B580?

The best overall model for the Intel Arc B580 is Qwen/Qwen3-8B-GGUF using the recommended Q8_0 quantization (9.47 GB total memory). With 12 GB of VRAM, this GPU typically runs a 14B model at Q5, comfortably at full GPU speed. Check our guide to the best LLMs for 12 GB VRAM for details.

Can the Intel Arc B580 run 14B models locally?

Yes. The Intel Arc B580's 12 GB VRAM is the exact sweet spot for running 14B models (such as Qwen 2.5 Coder 14B) in Q4_K_M quantization (~9.2 GB footprint) with full 4,096-token context entirely in VRAM.

How do you run local LLMs on an Intel Arc card like the Intel Arc B580?

Intel Arc GPUs run local AI models using the llama.cpp SYCL backend or Intel's IPEX-LLM library. Popular frontends like LM Studio and Ollama provide experimental or DirectML/Vulkan support for Arc hardware.