Llama 3.3 70B — Apple Silicon Benchmarks

Measured inference speed for Llama 3.3 70B across 1 Apple Silicon chip. Tokens per second at multiple quantization levels. Real runs, not estimates.

Quantizations measured: Q5_K_M

1 Benchmark rows

1 Chip tiers covered

7.1 Fastest avg tok/s (M4 Max (24-core GPU))

50 GB Minimum RAM observed

Benchmark results for Llama 3.3 70B

Rows sorted by avg tok/s descending. Click source badge to see original measurement page.

Chip	Quant	RAM req.	Context	Avg tok/s	Prompt tok/s	Runtime	Source
M4 Max (24-core GPU)	Q5_K_M	50.0 GB	8k	7.1 tok/s	—	—	ref

Chips with published results for Llama 3.3 70B

Data

benchmarks.json — full dataset · models.json — model summaries · benchmarks.csv — CSV export