What is the biggest model the Arc B580 can run?

Qwen3 14B, which needs about 11 GB, on the 12 GB version at about 27 tok/s.

What is the biggest model the Arc Pro B60 can run?

Qwen3 32B, which needs about 22 GB, on the 24 GB version at about 13 tok/s.

Arc B580 vs Arc Pro B60 for local AI

The Arc Pro B60 fits more: up to Qwen3 32B, against Qwen3 14B on the Arc B580. They generate at about the same speed, since their memory bandwidth is close (456 and 456 GB/s).

The numbers that matter

Memory and bandwidth

Memory decides which models fit. Bandwidth decides how fast they generate, because every token reads the model's active weights once.

	Arc B580	Arc Pro B60
Memory	12 GB VRAM	24 GB VRAM
Bandwidth	456 GB/s	456 GB/s
Biggest model that fits	Qwen3 14B	Qwen3 32B
Best everyday pick	Qwen3 14B, ~27 tok/s	Qwen3 30B-A3B, ~80 tok/s

Model by model

Every model, side by side

Estimated generation speed at the largest memory size each was sold in (12 GB and 24 GB).

Model	Arc B580	Arc Pro B60
Qwen3 4B About 4 GB	~100 tok/s Runs great	~100 tok/s Runs great
Qwen3 8B About 7 GB	~48 tok/s Runs great	~48 tok/s Runs great
Gemma 3 12B About 10 GB	~31 tok/s Runs great	~31 tok/s Runs great
Qwen3 14B About 11 GB	~27 tok/s Usable	~27 tok/s Usable
gpt-oss 20B About 16 GB	Needs 16 GB	~80 tok/s Runs great
Gemma 3 27B About 19 GB	Needs 19 GB	~15 tok/s Usable
Qwen3 30B-A3B About 21 GB	Needs 21 GB	~80 tok/s Runs great
Qwen3 32B About 22 GB	Needs 22 GB	~13 tok/s Usable
Llama 3.3 70B About 45 GB	Needs 45 GB	Needs 45 GB
gpt-oss 120B About 67 GB	Needs 67 GB	Needs 67 GB
Qwen3 235B-A22B About 144 GB	Needs 144 GB	Needs 144 GB
DeepSeek R1 671B About 406 GB	Needs 406 GB	Needs 406 GB

Which should you get?

It depends. The Arc Pro B60 fits bigger models; the Arc B580 runs the models both fit faster. Price matters as much as either number: check used prices on the builds page, where every parts list has a measured or estimated speed.

How we estimate

Both columns use the same formula: memory bandwidth divided by the weights read per token, calibrated on our own rack, with Ollama's default 4-bit sizes. Real results are usually within about a third. Measured either one? Send us the number.

Compare

Other matchups

Every comparison