LLM Inference on WebLLM macOS Apple M2 16GB unified memory
46.4Decode Throughput (tok/s)
Qwen2.5-0.5B
Updated 13d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.02 | 46.4 | 510 | |
2026.02 | 41.7 | 257 | |
2026.02 | 36 | 225 | |
2026.02 | 29.7 | 93 | |
2026.02 | 9.6 | 77 | |
2026.02 | 9.6 | 58 |
Qwen2.5-0.5B
| Method | Links | ||
|---|---|---|---|
2026.02 | 46.4 | 510 | |
2026.02 | 41.7 | 257 | |
2026.02 | 36 | 225 | |
2026.02 | 29.7 | 93 | |
2026.02 | 9.6 | 77 | |
2026.02 | 9.6 | 58 |