mlx-serve on the M4 Max 128GB
On the M4 Max 128GB, mlx-serve runs Qwen3.8 125B · 6B active at 8-bit at a median 60.0 tokens per second for generation, from 1 plain run (context not stated).
Speed by model on the M4 Max 128GB
Plain runs only: one device, one request, no speculative decoding, the whole model in memory. Each row is one model size at one bit class, with the other engines' plain runs of the same model on this card beside it. Context length and build still differ between runs, and each side shows its own.
| Model | mlx-serve | Other engines |
|---|---|---|
| Qwen3.8 125B · 6B active 8-bit | 60.0 t/sone report, context not stated | no other engine reported |
3 mlx-serve reports on the M4 Max 128GB
Newest first. A run on several cards, with speculative decoding, with part of the model in system RAM or with batched requests says so, since each describes a different setup.
- Qwen3.8 27B43.0 t/s(speculative decoding)
- Qwen3.8 125B · 6B active · mixed 4-8bit60.0 t/s
- Qwen3.8 125B · 6B active · Q473.8 t/s(speculative decoding)
Frequently asked
How fast is mlx-serve on the M4 Max 128GB?
On the M4 Max 128GB, mlx-serve runs Qwen3.8 125B · 6B active at 8-bit at a median 60.0 tokens per second for generation, from 1 plain run (context not stated).