Shared local-Mac total for 5Q, 20Q, and 10Q.
Hello: 1.8s · 20Q avg: 4.4s · 10Q avg: 8.4s
This page compares the local Apple-silicon Mac runs for qwen2.5-coder:0.5b, qwen2.5-coder:1.5b, and qwen2.5-coder:3b under the shared benchmark stack.
Shared local-Mac total for 5Q, 20Q, and 10Q.
Hello: 1.8s · 20Q avg: 4.4s · 10Q avg: 8.4s
Shared local-Mac total for 5Q, 20Q, and 10Q.
Hello: 4.8s · 20Q avg: 11.4s · 10Q avg: 24.5s
Shared local-Mac total for 5Q, 20Q, and 10Q.
Hello: 5.8s · 20Q avg: 13.3s · 10Q avg: 31.1s
This page compares only the local Apple-silicon Mac runs under the same one-model Ollama shape: one loaded model, one parallel slot, 4096 context, and the same benchmark packets. That makes it the cleanest view of how model size changes behavior on this Mac without the VPS in the loop.
| Metric | Qwen 0.5B | Qwen 1.5B | Qwen 3B |
|---|---|---|---|
| Total wall time | 88.0s | 68.6s | 110.1s |
| Average question time | 17.6s | 13.7s | 22.0s |
| Average throughput | 45.69 tok/s | 21.17 tok/s | 12.86 tok/s |
| Average marker hit | 80% | 83% | 90% |
| Format passes | 3/5 | 2/5 | 3/5 |
| Strict passes | 2/5 | 2/5 | 3/5 |
| Metric | Qwen 0.5B | Qwen 1.5B | Qwen 3B |
|---|---|---|---|
| Total wall time | 122.2s | 378.6s | 397.1s |
| Primary avg duration | 4.4s | 11.4s | 13.3s |
| Follow-up avg duration | 1.7s | 7.5s | 6.5s |
| Primary avg throughput | 42.18 tok/s | 16.89 tok/s | 14.20 tok/s |
| Primary avg marker hit | 78% | 88% | 85% |
| Usable primary answers | 20/20 | 20/20 | 20/20 |
| Metric | Qwen 0.5B | Qwen 1.5B | Qwen 3B |
|---|---|---|---|
| Total wall time | 104.3s | 322.2s | 407.3s |
| Primary avg duration | 8.4s | 24.5s | 31.1s |
| Follow-up avg duration | 2.0s | 7.7s | 9.6s |
| Primary avg throughput | 38.14 tok/s | 17.82 tok/s | 14.01 tok/s |
| Primary avg marker hit | 55% | 69% | 78% |
| Usable primary answers | 10/10 | 10/10 | 10/10 |
Every link below goes to a standalone archived report page.