Chat model by Google · 27.23B parameters
Everything this chip runs, machine by machine: MacBook Pro 14" M5 Pro, MacBook Pro 16" M5 Pro
txgemma 27b chat can run but will be tight at 75% of RAM. Consider a lower quantization or closing other apps.
~11 tok/s
Q4_K_M · 18.1 GB · Estimated
Alternative: Q3_K_M (15.4 GB, ~13 tok/s)
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 | 30.9 GB | ~6 tok/s | ✗ |
| Q6_K | 24.1 GB | ~8 tok/s | ✗ |
| Q5_K_M | 20.8 GB | ~9 tok/s | ✗ |
| Q4_K_M Recommended | 18.1 GB | ~11 tok/s | ✓ |
| Q3_K_M | 15.4 GB | ~13 tok/s | ✓ |
| Q2_K | 12.4 GB | ~17 tok/s | ✓ |
Closest matches by use case and parameter count.
ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.
Get ToolPiper — Free