Intel · Xe2 (Battlemage) · 2025

Intel Arc Pro B60 for local AI

24GB for about $650 new, among the lowest prices per GB of any current card, with less polished software than CUDA.

Updated October 14, 2026 · Speeds marked ~ are estimates; seemethodology

VRAM: 24GB
Bandwidth: 456 GB/s
Power: 200W
Runs well: 8 models

Arc Pro B60

Typically about $650

Can the Arc Pro B60 run LLMs?

The Arc Pro B60 has 24GB of GDDR6 and 456 GB/s of memory bandwidth, which decides how big a model fits and how fast it generates text. It comfortably runs models up to about 35B parameters: the largest in our catalog is Qwen3.6 35B-A3B at Q3_K_M (19GB with 8K context, ~66 tokens/sec). Our top all-round pick for it is Qwen3.8 27B (Q4_K_M, ~15 tokens/sec). Llama 3.3 70B Instruct only runs with partial CPU offload, at about 1.7 tokens/sec with CPU offload.

Recommendations

Best models for the Arc Pro B60

Full list

Best overall · also coding, reasoning, vision
Qwen3.8 27B
Q4_K_M · 18GB · ~15 tokens/sec
Runs well
Runner-up
Gemma 4 26B A4B
Q5_K_M · 20GB · ~43 tokens/sec
Runs well
Fastest good model · also largest that fits
Qwen3.6 35B-A3B
Q3_K_M · 19GB · ~66 tokens/sec
Runs well

Read: the best LLMs to run on the Arc Pro B60

Compatibility

What runs on the Arc Pro B60

Every current model in our catalog, at the best quantization that fits in 24GB with 8K tokens of context. "Partial offload" means part of the model runs from system RAM (we assume 64GB), which is much slower.

Which open-weight models run on the Intel Arc Pro B60, with the best quantization and estimated speed
Model	Parameters	Best quant	Memory	Speed	Verdict
Gemma 4 E4B	8B	Q8_0	10 GB	~25 tok/s	Runs well
Qwen3.5 9B	9B	Q8_0	12 GB	~22 tok/s	Runs well
Phi-4	14.7B	Q8_0	18 GB	~16 tok/s	Runs well
gpt-oss-20b	20.9B (3.6B active)	MXFP4	15 GB	~48 tok/s	Runs well
Gemma 4 26B A4B	26B (3.8B active)	Q5_K_M	20 GB	~43 tok/s	Runs well
Qwen3.8 27B	27B	Q4_K_M	18 GB	~15 tok/s	Runs well
Gemma 4 31B	31B	Q4_K_M	22 GB	~12 tok/s	Fits, tight
Qwen3.6 35B-A3B	35B (3B active)	Q3_K_M	19 GB	~66 tok/s	Runs well
GLM-4.5-Air	106B (12B active)	Q4_K_M	68 GB	~7.6 tok/s	Partial offload
Llama 4 Scout	109B (17B active)	Q4_K_M	70 GB	~5.4 tok/s	Partial offload
gpt-oss-120b	117B (5.1B active)	MXFP4	68 GB	~18 tok/s	Partial offload
Qwen3.5 122B-A10B	122B (10B active)	—	—	—	Won't fitWhat it needs
DeepSeek V4 Flash	284B (13B active)	—	—	—	Won't fitWhat it needs
GLM-5.3-Flash	320B (18B active)	—	—	—	Won't fitWhat it needs
GLM-5.3	744B (40B active)	—	—	—	Won't fitWhat it needs

Performance

Arc Pro B60 vs. alternatives on Qwen3.8 27B

Generation speed in tokens per second for a single user. Token generation is limited mostly by memory bandwidth, which is why bandwidth predicts these results so well.

Estimated (method)

Measured

Arc Pro B60 benchmark results

All benchmarks

Model	Quant	Runtime	Generation	Prompt	Source
Llama 2 7B Chat	Q4_0	llama.cpp (Vulkan)	69 tok/s	520 tok/s	llama.cpp discussion #10879: Performance of llama.cpp with Vulkan · October 2026

Our take

Should you buy the Arc Pro B60 for AI?

The Arc Pro B60 uses the same Battlemage chip as the Arc B580 but doubles memory to 24GB of GDDR6 on a 192-bit bus for 456 GB/s. Partner cards come in blower and single-fan designs rated from 120 to 200W, and some vendors sell dual-GPU boards with 48GB that need x8/x8 bifurcation on the motherboard. It runs 24B to 32B models at 4 bit through llama.cpp's SYCL and Vulkan backends and Intel's vLLM builds. New cards sold for about $650 in October 2026.

What we like

24GB for about $650 new
Low 120 to 200W power
Blower versions stack in multi-GPU builds

Watch out for

456 GB/s is slow for 24GB
Software support trails CUDA
Dual-GPU boards need bifurcation support

Best for: Budget 24GB and multi-GPU builds · Running 24B to 32B models at 4 bit

Specifications

Arc Pro B60 specs that matter for AI

Memory: 24GB GDDR6
Memory bandwidth: 456 GB/s
Memory bus: 192-bit
Architecture: Xe2 (Battlemage)
Power: 200W
Interface: PCIe 5.0 x8
Released: 2025
Street price (new): ~$650 (2026-10)
Software: SYCL, Vulkan

Sources: GPU Poet Arc Pro B60 specs and pricing, GPUvec Arc Pro B60 specifications

Cross-shop

Alternatives to the Arc Pro B60

All GPUs

Intel2026

32GB

GDDR6 ECC · 608 GB/s

Arc Pro B70

~$1,300

Intel2024

12GB

GDDR6 · 456 GB/s

Arc B580

~$330

NVIDIA2020

24GB

GDDR6X · 936 GB/s

RTX 3090

~$1,050 used

RTX 3090 vs Arc Pro B60

Frequently asked questions

How much VRAM does the Arc Pro B60 have?

The Intel Arc Pro B60 has 24GB of GDDR6 with 456 GB/s of memory bandwidth on a 192-bit bus.

What is the largest LLM the Arc Pro B60 can run?

Fully in GPU memory, the largest model in our catalog that runs comfortably is Qwen3.6 35B-A3B (35B (3B active)) at Q3_K_M, using about 19GB with 8K context at ~66 tokens/sec.

Can the Arc Pro B60 run Llama 3.3 70B?

Only with partial CPU offload. At Q4_K_M the model needs about 47GB with 8K context, so part of it spills into system RAM and speed drops to about 1.7 tokens/sec with CPU offload. For usable 70B speeds you want 48GB of VRAM or more (for example two 24GB cards).

What is the best LLM to run on the Arc Pro B60?

Our top pick is Qwen3.8 27B at Q4_K_M, which fits in 24GB with room for 8K of context and runs at ~15 tokens/sec. See our full list of the best models for the Arc Pro B60 for coding, reasoning, and vision picks.

Is the Arc Pro B60 good for local AI in 2026?

24GB for about $650 new, among the lowest prices per GB of any current card, with less polished software than CUDA.