← Back to community benchmarks
MiniMax-M2.5-REAP-139B-4bit
Performance
32k
tokens
332.3
PP tok/s
16.4
TG tok/s
98619
TTFT (ms)
77.5
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
96 GB
GPU Cores
60
Software
oMLX
v0.3.0rc1
macOS
macOS 26.4
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 568.1 | 48.6 | 69.8 GB | view |
| 4k | 602.1 | 41.0 | 70.8 GB | view |
| 8k | 550.2 | 33.9 | 71.7 GB | view |
| 16k | 458.8 | 26.2 | 73.7 GB | view |
| 32k | 332.3 | 16.4 | 77.5 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 48.6 | 1.00× |
| 2× | 70.4 | 1.45× |
| 4× | 87.5 | 1.80× |