inference/research
verifiedlow · 0.60updated 2026-10-15

M3 Max Qwen 3 30B-A3B Q4_K - Medium decode (Ollama, siliconscore ref)

28.0 tok/s (single-request decode, per-request)

Fields

id
benchmarks/m3max-qwen-3-30b-a3b-q4-k-medium-ollama-ctxnp-tg
Metric
decode_tok_s
Value
28
Unit
tok/s (single-request decode, per-request)
Model
Qwen 3 30B-A3B
Methodology
Silicon Score benchmarks.csv reference run. M3 Max (36 GB): Qwen 3 30B-A3B Q4_K - Medium via Ollama. context_tokens=not published. Decode tok/s single-request.
Measured by
third_party
Reproducible
yes
Thermal state
unknown
notes
siliconscore reference run; source_url=https://llmcheck.net/data/benchmarks.json. Chip variant 'M3 Max (36 GB)'.

Notes

siliconscore reference run; source_url=https://llmcheck.net/data/benchmarks.json. Chip variant 'M3 Max (36 GB)'.

Cross-references