← Back to community benchmarks
DeepSeek-R1-0528-Qwen3-8B
Performance
1k
tokens
215.9
PP tok/s
20.6
TG tok/s
4742
TTFT (ms)
4.9
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.3.8rc1
macOS
macOS 26.4.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 215.9 | 20.6 | 4.9 GB | current |
| 4k | 209.8 | 16.6 | 5.4 GB | view |
| 8k | 188.0 | 12.7 | 5.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 20.6 | 1.00× |
| 2× | 37.8 | 1.83× |
| 4× | 83.8 | 4.07× |
| 8× | 164.3 | 7.98× |