← Back to community benchmarks
Qwen-AgentWorld-35B-A3B
Performance
16k
tokens
2,427
PP tok/s
61.6
TG tok/s
6750
TTFT (ms)
66.7
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.5.1
macOS
macOS 26.5
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,415 | 65.2 | 65.6 GB | view |
| 4k | 2,813 | 64.2 | 66.3 GB | view |
| 8k | 2,709 | 61.8 | 66.5 GB | view |
| 16k | 2,427 | 61.6 | 66.7 GB | current |
| 32k | 2,083 | 57.6 | 67.4 GB | view |
| 64k | 1,594 | 50.2 | 68.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 65.2 | 1.00× |
| 2× | 73.6 | 1.13× |