Llama 3.1 8B Instruct › Apple A18 Pro (5-core GPU)

Llama 3.1 8B Instruct on the Apple A18 Pro (5-core GPU)

Best measured: 15.8 tokens/s generation and 116 tokens/s prompt processing (Q4_K_M, llamafile).

Every run

Quant	Backend	Generate tok/s	Prompt tok/s	Test	Build / by / notes	Source
Q4_K_M	llamafile	15.8	116	LocalScore avg	LocalScore rank 180/422	LocalScore (localscore.ai)

tg128 = generating 128 tokens; pp512 = reading a 512-token prompt (llama-bench). LocalScore averages several prompt and output lengths, so its numbers run lower than tg128.

Check prices

Amazon

As an Amazon Associate I earn from qualifying purchases.

No card yet?

Try it on a rented GPU (RunPod)

Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.

All results for the Apple A18 Pro (5-core GPU) · Llama 3.1 8B Instruct on other hardware