← CanIRunLocal · All hardware · All models
Prism ML · 27B parameters · chat, vision, reasoning.
Ollama weight files: Q2_0 8.3 GB (bonsai-2:27b).
Yes on 53 machines in this list. The smallest is RTX 3060 Ti 8GB: Q2_0, about 10.6 GB, roughly 12 tokens per second, via Ollama. ollama run bonsai-2:27b
ollama run bonsai-2:27bollama run bonsai-2:27bollama run bonsai-2:27bollama run bonsai-2:27b| Hardware | Runtime | Quant | Memory | Fit | Speed | Command |
|---|---|---|---|---|---|---|
| RTX 3060 Ti 8GB | Ollama | Q2_0 | 10.6 GB | partial offload (slower) | ~12 tok/s | ollama run bonsai-2:27b |
| RTX 3070 8GB | Ollama | Q2_0 | 10.6 GB | partial offload (slower) | ~12 tok/s | ollama run bonsai-2:27b |
| RTX 4060 8GB | Ollama | Q2_0 | 10.6 GB | partial offload (slower) | ~10 tok/s | ollama run bonsai-2:27b |
| RX 7600 8GB | Ollama | Q2_0 | 10.6 GB | partial offload (slower) | ~11 tok/s | ollama run bonsai-2:27b |
| RTX 3080 10GB | Ollama | Q2_0 | 10.6 GB | partial offload (slower) | ~26 tok/s | ollama run bonsai-2:27b |
| M1 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~7 tok/s | ollama run bonsai-2:27b |
| M2 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~10 tok/s | ollama run bonsai-2:27b |
| M3 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~10 tok/s | ollama run bonsai-2:27b |
| M4 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~12 tok/s | ollama run bonsai-2:27b |
| RTX 3060 12GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~26 tok/s | ollama run bonsai-2:27b |
| RTX 4070 12GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~36 tok/s | ollama run bonsai-2:27b |
| RTX 5070 12GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~49 tok/s | ollama run bonsai-2:27b |
| RTX 4060 Ti 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~21 tok/s | ollama run bonsai-2:27b |
| RTX 4070 Ti Super 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~49 tok/s | ollama run bonsai-2:27b |
| RTX 4080 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~52 tok/s | ollama run bonsai-2:27b |
| RTX 5080 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~69 tok/s | ollama run bonsai-2:27b |
| RX 7800 XT 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~45 tok/s | ollama run bonsai-2:27b |
| RX 9070 XT 16GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~46 tok/s | ollama run bonsai-2:27b |
| M2 24GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~10 tok/s | ollama run bonsai-2:27b |
| M3 24GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~10 tok/s | ollama run bonsai-2:27b |
| M4 24GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~12 tok/s | ollama run bonsai-2:27b |
| RTX 3090 24GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~68 tok/s | ollama run bonsai-2:27b |
| RTX 4090 24GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~73 tok/s | ollama run bonsai-2:27b |
| RX 7900 XTX 24GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~69 tok/s | ollama run bonsai-2:27b |
| M1 Pro 32GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~19 tok/s | ollama run bonsai-2:27b |
| M2 Pro 32GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~19 tok/s | ollama run bonsai-2:27b |
| M4 32GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~12 tok/s | ollama run bonsai-2:27b |
| M3 Pro 36GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~14 tok/s | ollama run bonsai-2:27b |
| Mac Studio M4 Max (32c GPU) 36GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~40 tok/s | ollama run bonsai-2:27b |
| Mac Studio M5 Max (32c GPU) 36GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~44 tok/s | ollama run bonsai-2:27b |
| RTX 5090 32GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~130 tok/s | ollama run bonsai-2:27b |
| M4 Pro 48GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~26 tok/s | ollama run bonsai-2:27b |
| Mac Studio M4 Max (40c GPU) 48GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~53 tok/s | ollama run bonsai-2:27b |
| Mac Studio M5 Max (40c GPU) 48GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~59 tok/s | ollama run bonsai-2:27b |
| M1 Max 64GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~39 tok/s | ollama run bonsai-2:27b |
| M2 Max 64GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~39 tok/s | ollama run bonsai-2:27b |
| M3 Max 64GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~39 tok/s | ollama run bonsai-2:27b |
| M4 Max 64GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~53 tok/s | ollama run bonsai-2:27b |
| Mac Studio M4 Max (40c GPU) 64GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~53 tok/s | ollama run bonsai-2:27b |
| Mac Studio M5 Max (40c GPU) 64GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~59 tok/s | ollama run bonsai-2:27b |
| DGX Spark 64GB (ASUS, Dell, HP, MSI, Acer, Gigabyte) | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~25 tok/s | ollama run bonsai-2:27b |
| Mac Studio M3 Ultra 96GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~79 tok/s | ollama run bonsai-2:27b |
| Mac Studio M5 Ultra 96GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~116 tok/s | ollama run bonsai-2:27b |
| M3 Max 128GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~39 tok/s | ollama run bonsai-2:27b |
| M4 Max 128GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~53 tok/s | ollama run bonsai-2:27b |
| Mac Studio M4 Max (40c GPU) 128GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~53 tok/s | ollama run bonsai-2:27b |
| Mac Studio M5 Max (40c GPU) 128GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~59 tok/s | ollama run bonsai-2:27b |
| DGX Spark 128GB (NVIDIA, ASUS, Dell, Lenovo, HP, MSI, Acer, Gigabyte) | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~25 tok/s | ollama run bonsai-2:27b |
| M2 Ultra 192GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~77 tok/s | ollama run bonsai-2:27b |
| Mac Studio M3 Ultra 256GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~79 tok/s | ollama run bonsai-2:27b |
| Mac Studio M5 Ultra (64c GPU) 256GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~116 tok/s | ollama run bonsai-2:27b |
| Mac Studio M5 Ultra (80c GPU) 256GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~116 tok/s | ollama run bonsai-2:27b |
| Mac Studio M3 Ultra (80c GPU) 512GB | Ollama | Q2_0 | 10.6 GB | fits in fast memory | ~79 tok/s | ollama run bonsai-2:27b |
Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-18.
A project by OnDevice Labs · llms.txt