Builds / Apple M5 Pro (unified, Mac mini)
Mac mini M5 Pro 64GB (MLX 4-bit): Qwen3.6 35B-A3B at 100.2 tok/s, Gemma 4 26B-A4B at 85.1, Qwen3.8 27B at 17.7 (mean over 1,000 rounds)
Last updated:
- GPU(s)
- Apple M5 Pro (unified, Mac mini)
- VRAM / unified memory (GB)
- 64
- CPU
- Apple M5 Pro
- Motherboard
- not stated
- RAM
- 64 GB unified memory
- PSU
- not stated
- Other
- macOS 27.0.1; Python 3.12; mlx-lm in-process; stock GPU wired-memory limit (not raised)
- Reported total cost
- not stated
- Power
- not stated
Measured speed (generation tokens/sec)
| Model | Quant | Engine | tok/s | Prompt processing (tok/s) | Details |
|---|---|---|---|---|---|
| Qwen3.6 35B-A3B | MLX 4-bit (mlx-community/Qwen3.6-35B-A3B-4bit) | mlx-lm (MLX 0.32.x) | 100.24 | not stated | context: short prompt (~250-410 tokens), 1,000 consecutive rounds📏 1,000 Rewrites Later dataset (CC BY 4.0): the model rewrote its own previous output 1,000 times in a row (Gettysburg Address seed, ~250-410 prompt tokens, up to ~360 output tokens per round, temp 0.7, top_p 0.95, seed 42, thinking off); mlx-lm (MLX 0.32.x) in-process, weights fully in memory, stock wired-memory limit; listed figure is the mean generation tok/s over all 1,000 rounds as reported by the runtime (dataset summary), measured 23 Oct 2026; median 100.41 tok/s; MLX peak memory 20.15 GB; 90.19 mean in a second 1,000-round run at temp 1.0 |
| Gemma 4 26B-A4B | MLX 4-bit (mlx-community/gemma-4-26b-a4b-it-4bit) | mlx-lm (MLX 0.32.x) | 85.08 | not stated | context: short prompt (~250-410 tokens), 1,000 consecutive rounds📏 1,000 Rewrites Later dataset (CC BY 4.0): the model rewrote its own previous output 1,000 times in a row (Gettysburg Address seed, ~250-410 prompt tokens, up to ~360 output tokens per round, temp 0.7, top_p 0.95, seed 42, thinking off); mlx-lm (MLX 0.32.x) in-process, weights fully in memory, stock wired-memory limit; listed figure is the mean generation tok/s over all 1,000 rounds as reported by the runtime (dataset summary), measured 23 Oct 2026; median 85.47 tok/s; MLX peak memory 14.69 GB; 84.1 mean in a second 1,000-round run at temp 1.0 |
| Qwen3.8 27B | MLX 4-bit (mlx-community/Qwen3.8-27B-4bit) | mlx-lm (MLX 0.32.x) | 17.7 | not stated | context: short prompt (~250-410 tokens), 1,000 consecutive rounds📏 1,000 Rewrites Later dataset (CC BY 4.0): the model rewrote its own previous output 1,000 times in a row (Gettysburg Address seed, ~250-410 prompt tokens, up to ~360 output tokens per round, temp 0.7, top_p 0.95, seed 42, thinking off); mlx-lm (MLX 0.32.x) in-process, weights fully in memory, stock wired-memory limit; listed figure is the mean generation tok/s over all 1,000 rounds as reported by the runtime (dataset summary), measured 23 Oct 2026; median 17.81 tok/s; MLX peak memory 16.15 GB; 3.0 h of generation, 180,209 tokens; per-round prefill was ~350-405 tok/s on these short prompts |
Dataset collected 23-25 Oct 2026 on one Mac mini (Apple M5 Pro, 64 GB) and published on Hugging Face under CC BY 4.0, with every round's speed in the per-round files. Because each figure is the mean over 1,000 back-to-back generations, it is a sustained speed rather than a peak. Not listed: Qwen3-Next 80B-A3B 4-bit (71.4 tok/s), GLM-4.7 Flash 4-bit (83.7), DeepSeek-V2-Lite 4-bit (136.1) and Qwen3.5 0.8B 4-bit (349.8), which are not yet RigBase models; Bonsai 2 9B (2-bit ternary) and the pruned mixed-quant Qwen3.8 Flash-Next (below Q4 / not a stock quant); and the GLM-5.3 Flash (2 bpw) and DeepSeek V4.1 Flash (~2.5-bit) runs streamed from the SSD (5.8 and 5.5 tok/s), which are below Q4. The same author's 17.8 tok/s short-prompt figure on this machine is also posted in the mlx-community Qwen3.8-27B-4bit discussion #4.
Blank fields = not stated in the source. We never estimate build data; GPU street prices above are separately sourced and dated.
Copy into a README or blog. Links back to this build. See badge docs (CC BY 4.0 attribution). The SVG does not invent speeds — numbers stay on the build page.
Markdown
[](https://getrigbase.com/builds/mac-mini-m5-pro-64gb-mlx-1000-rewrites-pratikgx/)
HTML
<a href="https://getrigbase.com/builds/mac-mini-m5-pro-64gb-mlx-1000-rewrites-pratikgx/" rel="noopener"><img src="https://getrigbase.com/badge/mac-mini-m5-pro-64gb-mlx-1000-rewrites-pratikgx.svg" alt="Measured on RigBase" height="20"/></a>
Preview: 
· How to use