← Back to community benchmarks
GLM-4.7-Flash
Performance
32k
tokens
628.5
PP tok/s
33.5
TG tok/s
52139
TTFT (ms)
32.9
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.6
macOS
macOS 26.3
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,310 | 64.1 | 23.3 GB | view |
| 4k | 1,627 | 58.6 | 24.3 GB | view |
| 8k | 1,397 | 53.2 | 25.5 GB | view |
| 16k | 1,005 | 43.8 | 27.9 GB | view |
| 32k | 628.5 | 33.5 | 32.9 GB | current |
| 64k | 346.7 | 20.9 | 43.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 64.1 | 1.00× |
| 2× | 103.0 | 1.61× |
| 4× | 160.1 | 2.50× |
| 8× | 209.9 | 3.27× |