Apple M4 Max (128GB, 40-core) Llama-3.3-70B 4-bit decode
11.8 tok/s (4-bit); 8-bit 6.5
Fields
id
benchmarks/comm-apple-m4-max-llama33-70b-4bit-decodeMetric
decode_tok_s
Value
11.8
Unit
tok/s (4-bit); 8-bit 6.5
Model
Llama-3.3-70B 4-bit
Methodology
Silicon Score measured; 128GB MacBook Pro, ~546 GB/s
Measured by
third_party
Thermal state
unknown