Leaderboard/qwen3-coder-bench-32k/M4 Max (40-core GPU)/r_roktphpc--8

qwen3-coder-bench-32k on M4 Max (40-core GPU) + 128GB unified

Name: qwen3-coder-bench-32k on M4 Max (40-core GPU) + 128GB unified
Creator: llm-speed
License: https://www.apache.org/licenses/LICENSE-2.0
Keywords: LLM benchmark, tokens per second, qwen3-coder-bench-32k, M4 Max (40-core GPU) + 128GB unified, ollama

M4 Max (40-core GPU) + 128GB unified

suite suite-v1

cli 0.0.3

signedya0OvDcfH4…

Embed badgesubmitted Jun 27, 2026

https://llm-speed.com/r/r_roktphpc--8

Workload results

Workload	Backend	Model	decode tok/s	prefill tok/s	TTFT	p50	p95
chat-short	ollama@0.30.11	qwen3-coder-bench-32kQ4_K_M	113.3tok/s	84.12tok/s	1,308ms	8.9ms	9.2ms
chat-long	ollama@0.30.11	qwen3-coder-bench-32kQ4_K_M	97.40tok/s	1,342.8tok/s	2,344ms	10.3ms	10.8ms
concurrent-decode	ollama@0.30.11	qwen3-coder-bench-32kQ4_K_M	109.7tok/s	—	—	9.1ms	9.5ms
agent-trace	ollama@0.30.11	qwen3-coder-bench-32kQ4_K_M	103.2tok/s	3,371.6tok/s	477ms	9.5ms	10.5ms

Reproduce on your machine

Same workload, same model, signed at your rig. The exact command that produced this run:

$ pipx install llm-speed && llm-speed bench --model 'qwen3-coder-bench-32k' --workload 'chat-short'

Runs in about a minute. Your number lands on the leaderboard signed and linkable. How it's measured.

Embed this run

Drop the badge into a README, blog post, or signature. Each render is a backlink to the signed result.

[![llm-speed: 113 tok/s on M4 Max (40-core GPU) (qwen3-coder-bench-32k)](https://llm-speed.com/badge/r_roktphpc--8.svg)](https://llm-speed.com/r/r_roktphpc--8)

Related benchmarks

More qwen3-coder-bench-32k benchmarks — every backend and rig that has run this model.
More M4 Max (40-core GPU) LLM benchmarks — every model measured on this hardware.

Provenance

Run ID: r_roktphpc--8
Fingerprint hash: 275ecb2b79296aab
Public key: ya0OvDcfH4La0mEhEhM8iESwvi9/MZz9uibPMNfpovE=
Received: 2026-06-27 16:14:03