← Back to community benchmarks
Qwen3.5-9B-Claude-4.6-HighIQ-INSTRUCT-HERETIC-UNCENSOREDmxfp8
Performance
1k
tokens
302.3
PP tok/s
36.5
TG tok/s
3387
TTFT (ms)
9.7
Peak mem (GB)
Hardware
Chip
M1 Max (32c)
Memory
64 GB
GPU Cores
32
Software
oMLX
v0.3.4
macOS
macOS 26.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 302.3 | 36.5 | 9.7 GB | current |
| 4k | 323.2 | 35.6 | 10.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 36.5 | 1.00× |
| 2× | 47.6 | 1.30× |
| 4× | 51.1 | 1.40× |