← Back to community benchmarks
Qwen2.5-VL-7B-Instruct
Performance
4k
tokens
217.1
PP tok/s
11.7
TG tok/s
18866
TTFT (ms)
9.2
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 236.8 | 12.2 | 8.9 GB | view |
| 4k | 217.1 | 11.7 | 9.2 GB | current |
| 8k | 209.3 | 11.7 | 9.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 12.2 | 1.00× |
| 2× | 24.2 | 1.98× |
| 4× | 32.3 | 2.65× |
| 8× | 38.8 | 3.18× |