M3 Max 96GB vs M5 Max 128GB for local LLMs
2 reports on the M3 Max 96GB and 48 on the M5 Max 128GB
Which is faster for local LLMs?
On Qwen3.6 27B at 8-bit, the M3 Max 96GB has 7.4 generation tokens/s (one run); prompt processing 121.3 tokens/s (one run); the M5 Max 128GB has 17.2 generation tokens/s (one run); prompt processing 433.0 tokens/s (2 runs). Those are community reports whose engines and settings differ, so the gap is a rough sign, not a controlled measure of the two cards. Memory bandwidth, which caps how fast a card can write, is 2.0x higher on the M5 Max 128GB (307 GB/s on the M3 Max 96GB, 614 on the M5 Max 128GB).
- Memory
- 96GB unified
- Memory bandwidth
- 307 GB/s
- FP16 compute
- 10.6 TFLOPS
- Reports
- 2
- Memory
- 128GB unified
- Memory bandwidth
- 614 GB/s
- FP16 compute
- not on record
- Reports
- 48
Measured on both
Generation and prompt processing speeds from plain runs (one device, one request, no speculative decoding, the whole model in the device's memory) of the same model size at the same bit class. Each side shows the engines and reported context behind its median. Context can mean a configured window or prompt depth, so these are community comparisons with differing settings. How to read these
| Model | M3 Max 96GB | M5 Max 128GB |
|---|---|---|
| Qwen3.6 27B8-bit | 7.4one runPP 121.3 (one run)oMLXcontext not stated | 17.2one runPP 433.0 (2 runs)MLX125K to 290K context |
Estimated by the calculator
The quant the calculator recommends for each card and its estimated speed at a 32,768-token context with 32 GB of system RAM, from memory bandwidth and the model's shape. Where a model was also measured above, the measurement wins.
| Model | M3 Max 96GB | M5 Max 128GB |
|---|---|---|
| Qwen3.6 27B | 4.3 t/sFP16 | 9.0 t/sFP16 |
| Qwen3.8 27B | 4.3 t/sFP16 | 9.0 t/sFP16 |
| Qwen3.8 125B · 6B active | 29.7 t/sQ3_K_M | 41.2 t/sQ5_K_S |
| Qwen3.6 35B · 3B active | 30.8 t/sQ8_0 | 28.5 t/sFP16 |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | 64.8 t/sQ2_K |
| Gemma 4 26B · 4B active | 10.2 t/sFP16 | 24.4 t/sFP16 |
| DeepSeek V4.1 Flash 552B · 16B active | doesn't fit at these settings | doesn't fit at these settings |
Change the context, RAM or model in the calculator: M3 Max 96GB · M5 Max 128GB
Every report on each card, with its source: M3 Max 96GB · M5 Max 128GB