M1 Max 64GB vs M2 Ultra 128GB for local LLMs
45 reports on the M1 Max 64GB and 4 on the M2 Ultra 128GB
Which is faster for local LLMs?
Nobody has reported the same model at the same quant on both cards yet, so there's no measured head-to-head. Memory bandwidth, which caps how fast a card can write, is 2.0x higher on the M2 Ultra 128GB (410 GB/s on the M1 Max 64GB, 819 on the M2 Ultra 128GB). The estimates below compare them on the models people run most.
M1 Max 64GB
- Memory
- 64GB unified
- Memory bandwidth
- 410 GB/s
- FP16 compute
- 10.4 TFLOPS
- Reports
- 45
M2 Ultra 128GB
- Memory
- 128GB unified
- Memory bandwidth
- 819 GB/s
- FP16 compute
- 27.2 TFLOPS
- Reports
- 4
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.
| Model | M1 Max 64GB | M2 Ultra 128GB |
|---|---|---|
| Qwen3.8 27B | 10.0 t/sQ8_0 | 8.3 t/sFP16 |
| Qwen3.8 125B · 6B active | 45.7 t/sQ2_K | 37.9 t/sQ5_K_S |
| Qwen3.6 35B · 3B active | 31.6 t/sQ8_0 | 26.2 t/sFP16 |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | 59.6 t/sQ2_K |
| Qwen3.6 27B | 10.0 t/sQ8_0 | 8.3 t/sFP16 |
| Gemma 4 26B · 4B active | 27.1 t/sQ8_0 | 22.5 t/sFP16 |
Change the context, RAM or model in the calculator: M1 Max 64GB · M2 Ultra 128GB
Every report on each card, with its source: M1 Max 64GB · M2 Ultra 128GB