Unified MemoryBenchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Models
Ling-3.0-tiny (7.9B-A1.3B)
inclusionAI/Ling-3.0-tiny-GGUF:Q4_K_M · llama.cpp · llama-server version: 0.7.0 (build 11669, commit 620d2225c)
#6 of 8
overall, text jobs
80.5
quality score
117.9
median tokens/s
5.17 GB
RAM with 8K context filled
4.49 GB
download size
Score per job
| Job | Accuracy | Pass | Easy | Medium | Hard | Median latency | Tok/s | Job time | Errors |
|---|---|---|---|---|---|---|---|---|---|
| Assistant actions | 80.4 | 74.0% | 86.7% | 64.7% | 72.2% | 0.45 s | 113.2 | 26.8 s | 0 |
| Ask my documents | 76.6 | 70.0% | 66.7% | 69.2% | 73.3% | 1.09 s | 111.0 | 45.1 s | 0 |
| Rewrite and proofread | 77.7 | 26.2% | 21.4% | 35.7% | 21.4% | 0.84 s | 117.9 | 37.5 s | 0 |
| Summarize + extract | 87.1 | 47.7% | 53.3% | 46.2% | 43.8% | 1.07 s | 116.3 | 48.4 s | 0 |
| Translate | 84.5 | 70.0% | 76.9% | 76.9% | 57.1% | 0.68 s | 118.4 | 26.4 s | 0 |
| Sort and triage | 76.8 | 73.2% | 82.4% | 73.7% | 65.0% | 0.24 s | 120.4 | 12.5 s | 0 |
Easy / medium / hard columns are pass rates within each difficulty tier.
Memory and loading
| Download size | 4.49 GB |
| Loaded, idle | 4.97 GB |
| In use (8K context) | 5.07 GB |
| 8K context filled | 5.17 GB |
| Server peak during the jobs (incl. reserved context) | 13.59 GB |
| Load time | 1m 43s |
Measured on Apple M4 Max, 36 GB, macOS 27.0.1. How memory was measured.