SiliconScore community row: Apple M3 Ultra, MLX, Qwen3-Coder-Next 4bit MLX, decode 74.0 t/s
74.0 tok/s avg decode (single-request)
Fields
id
benchmarks/ss-m3-ultra-qwen3-coder-next-4bit-decodeMetric
decode_tok_s
Value
74
Unit
tok/s avg decode (single-request)
Model
Qwen3-Coder-Next (4bit MLX)
Methodology
COMMUNITY-UNVERIFIED row aggregated by siliconscore.com M3 Ultra chip page (accessed 2026-10-15); underlying evidence is the linked r/LocalLLaMA/HF measurement. MLX runtime, 4-bit-ish quant as named. avg decode tok/s single-request. Low-confidence forum-class evidence. RAM tier: 256 GB.
Measured by
third_party
Reproducible
no