← Back to community benchmarks
Llama-3.3-70B-Instruct
Performance
1k
tokens
50.5
PP tok/s
7.5
TG tok/s
20300
TTFT (ms)
29.6
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.3.6
macOS
macOS 26.3.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 50.5 | 7.5 | 29.6 GB | current |
| 4k | 45.3 | 7.0 | 30.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 7.5 | 1.00× |
| 2× | 8.8 | 1.17× |
| 4× | 9.2 | 1.23× |