Unified MemoryBenchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Models
MiniCPM5 2B
openbmb/MiniCPM5-2B-MLX · MLX · mlx_lm.server 0.32.2
#7 of 8
overall, text jobs
77.8
quality score
99.3
median tokens/s
2.36 GB
RAM with 8K context filled
1.32 GB
download size
Score per job
| Job | Accuracy | Pass | Easy | Medium | Hard | Median latency | Tok/s | Job time | Errors |
|---|---|---|---|---|---|---|---|---|---|
| Assistant actions | 81.7 | 74.0% | 93.3% | 70.6% | 61.1% | 0.46 s | 83.6 | 27.1 s | 0 |
| Ask my documents | 81.7 | 75.0% | 100.0% | 84.6% | 46.7% | 1.38 s | 103.6 | 56.1 s | 0 |
| Rewrite and proofread | 67.6 | 21.4% | 28.6% | 35.7% | 0.0% | 1.23 s | 81.2 | 54.5 s | 0 |
| Summarize + extract | 88.8 | 47.7% | 60.0% | 23.1% | 56.2% | 1.32 s | 98.3 | 58.5 s | 0 |
| Translate | 74.7 | 42.5% | 46.2% | 46.2% | 35.7% | 0.80 s | 99.3 | 30.0 s | 0 |
| Sort and triage | 72.3 | 66.1% | 76.5% | 63.2% | 60.0% | 0.34 s | 100.2 | 18.4 s | 0 |
Easy / medium / hard columns are pass rates within each difficulty tier.
Memory and loading
| Download size | 1.32 GB |
| Loaded, idle | 1.51 GB |
| In use (8K context) | 1.64 GB |
| 8K context filled | 2.36 GB |
| Server peak during the jobs (incl. reserved context) | 3.57 GB |
| Load time | 36.4 s |
Measured on Apple M4 Max, 36 GB, macOS 27.0.1. How memory was measured.