← CanIRunLocal · All hardware · All models
Kev · 4B parameters · chat.
Ollama weight files: Q4_K_M 2.7 GB (kev:4b-q4_K_M); Q8_0 4.5 GB (kev:4b-q8_0).
Yes on 85 machines in this list. The smallest is M1 8GB: Q8_0, about 5.3 GB, roughly 12 tokens per second, via Ollama. ollama run kev:4b-q8_0
ollama run kev:4b-q8_0ollama run kev:4b-q8_0ollama run kev:4b-q8_0ollama run kev:4b-q8_0lms get ggml-org/Kev-4B-GGUF@Q8_0llama-server -hf ggml-org/Kev-4B-GGUF:Q8_0 -c 8192| Hardware | Runtime | Quant | Memory | Fit | Speed | Command |
|---|---|---|---|---|---|---|
| M1 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~12 tok/s | ollama run kev:4b-q8_0 |
| M2 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~18 tok/s | ollama run kev:4b-q8_0 |
| M3 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~18 tok/s | ollama run kev:4b-q8_0 |
| RTX 3060 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~32 tok/s | ollama run kev:4b-q8_0 |
| RTX 3060 Ti 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~60 tok/s | ollama run kev:4b-q8_0 |
| RTX 3070 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~60 tok/s | ollama run kev:4b-q8_0 |
| RTX 4060 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~36 tok/s | ollama run kev:4b-q8_0 |
| RTX 4060 Ti 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~38 tok/s | ollama run kev:4b-q8_0 |
| RX 7600 8GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~38 tok/s | ollama run kev:4b-q8_0 |
| RTX 3080 10GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~101 tok/s | ollama run kev:4b-q8_0 |
| M1 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~12 tok/s | ollama run kev:4b-q8_0 |
| M1 Pro 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~36 tok/s | ollama run kev:4b-q8_0 |
| M2 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~18 tok/s | ollama run kev:4b-q8_0 |
| M2 Pro 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~36 tok/s | ollama run kev:4b-q8_0 |
| M3 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~18 tok/s | ollama run kev:4b-q8_0 |
| M4 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~21 tok/s | ollama run kev:4b-q8_0 |
| RTX 3060 12GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~48 tok/s | ollama run kev:4b-q8_0 |
| RTX 3080 12GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~122 tok/s | ollama run kev:4b-q8_0 |
| RTX 4070 12GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~67 tok/s | ollama run kev:4b-q8_0 |
| RTX 5070 12GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~90 tok/s | ollama run kev:4b-q8_0 |
| M3 Pro 18GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~27 tok/s | ollama run kev:4b-q8_0 |
| RTX 4060 Ti 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~38 tok/s | ollama run kev:4b-q8_0 |
| RTX 4070 Ti Super 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~90 tok/s | ollama run kev:4b-q8_0 |
| RTX 4080 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~96 tok/s | ollama run kev:4b-q8_0 |
| RTX 5080 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~128 tok/s | ollama run kev:4b-q8_0 |
| RX 7800 XT 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~83 tok/s | ollama run kev:4b-q8_0 |
| RX 9070 XT 16GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~85 tok/s | ollama run kev:4b-q8_0 |
| M2 24GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~18 tok/s | ollama run kev:4b-q8_0 |
| M3 24GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~18 tok/s | ollama run kev:4b-q8_0 |
| M4 24GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~21 tok/s | ollama run kev:4b-q8_0 |
| M4 Pro 24GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~49 tok/s | ollama run kev:4b-q8_0 |
| RTX 3090 24GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~125 tok/s | ollama run kev:4b-q8_0 |
| RTX 4090 24GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~134 tok/s | ollama run kev:4b-q8_0 |
| RX 7900 XTX 24GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~128 tok/s | ollama run kev:4b-q8_0 |
| M1 Max (24c GPU) 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M1 Max (32c GPU) 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M1 Pro 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~36 tok/s | ollama run kev:4b-q8_0 |
| M2 Max (30c GPU) 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M2 Max (38c GPU) 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M2 Pro 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~36 tok/s | ollama run kev:4b-q8_0 |
| M4 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~21 tok/s | ollama run kev:4b-q8_0 |
| M3 Max (30c GPU) 36GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~53 tok/s | ollama run kev:4b-q8_0 |
| M3 Pro 36GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~27 tok/s | ollama run kev:4b-q8_0 |
| M4 Max (32c GPU) 36GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~73 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M4 Max (32c GPU) 36GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~73 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Max (32c GPU) 36GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~82 tok/s | ollama run kev:4b-q8_0 |
| RTX 5090 32GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~239 tok/s | ollama run kev:4b-q8_0 |
| M3 Max (40c GPU) 48GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M4 Max (40c GPU) 48GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~97 tok/s | ollama run kev:4b-q8_0 |
| M4 Pro 48GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~49 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M4 Max (40c GPU) 48GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~97 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Max (40c GPU) 48GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~109 tok/s | ollama run kev:4b-q8_0 |
| M1 Max (32c GPU) 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M2 Max (30c GPU) 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M2 Max (38c GPU) 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M2 Ultra 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~142 tok/s | ollama run kev:4b-q8_0 |
| M3 Max (40c GPU) 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M4 Max (40c GPU) 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~97 tok/s | ollama run kev:4b-q8_0 |
| M4 Pro 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~49 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M4 Max (40c GPU) 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~97 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Max (40c GPU) 64GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~109 tok/s | ollama run kev:4b-q8_0 |
| M2 Max (38c GPU) 96GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M3 Max (30c GPU) 96GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~53 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M3 Ultra (60c GPU) 96GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~146 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M3 Ultra (80c GPU) 96GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~146 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Ultra (64c GPU) 96GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~213 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Ultra (80c GPU) 96GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~213 tok/s | ollama run kev:4b-q8_0 |
| M2 Ultra 128GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~142 tok/s | ollama run kev:4b-q8_0 |
| M3 Max (40c GPU) 128GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~71 tok/s | ollama run kev:4b-q8_0 |
| M4 Max (40c GPU) 128GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~97 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M4 Max (40c GPU) 128GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~97 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Max (40c GPU) 128GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~109 tok/s | ollama run kev:4b-q8_0 |
| DGX Spark 128GB (NVIDIA, ASUS, Dell, Lenovo, HP, MSI, Acer, Gigabyte) | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~46 tok/s | ollama run kev:4b-q8_0 |
| M2 Ultra 192GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~142 tok/s | ollama run kev:4b-q8_0 |
| 2x Mac Studio M5 Max (40c GPU) 128GB | llama.cpp | Q8_0 GGUF | 5.3 GB | fits in fast memory | ~93 tok/s | llama-server -hf ggml-org/Kev-4B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| Mac Studio M3 Ultra (60c GPU) 256GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~146 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M3 Ultra (80c GPU) 256GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~146 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Ultra (64c GPU) 256GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~213 tok/s | ollama run kev:4b-q8_0 |
| Mac Studio M5 Ultra (80c GPU) 256GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~213 tok/s | ollama run kev:4b-q8_0 |
| 2x DGX Spark 128GB | llama.cpp | Q8_0 GGUF | 5.3 GB | fits in fast memory | ~39 tok/s | llama-server -hf ggml-org/Kev-4B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| 4x Mac Studio M5 Max (40c GPU) 128GB | llama.cpp | Q8_0 GGUF | 5.3 GB | fits in fast memory | ~82 tok/s | llama-server -hf ggml-org/Kev-4B-GGUF:Q8_0 --rpc node2:50052,node3:50052,node4:50052 -c 8192 # run rpc-server on each other node |
| 2x Mac Studio M5 Ultra (80c GPU) 256GB | llama.cpp | Q8_0 GGUF | 5.3 GB | fits in fast memory | ~182 tok/s | llama-server -hf ggml-org/Kev-4B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| Mac Studio M3 Ultra (80c GPU) 512GB | Ollama | Q8_0 | 5.3 GB | fits in fast memory | ~146 tok/s | ollama run kev:4b-q8_0 |
| 4x DGX Spark 128GB | llama.cpp | Q8_0 GGUF | 5.3 GB | fits in fast memory | ~34 tok/s | llama-server -hf ggml-org/Kev-4B-GGUF:Q8_0 --rpc node2:50052,node3:50052,node4:50052 -c 8192 # run rpc-server on each other node |
| 4x Mac Studio M5 Ultra (80c GPU) 256GB | llama.cpp | Q8_0 GGUF | 5.3 GB | fits in fast memory | ~161 tok/s | llama-server -hf ggml-org/Kev-4B-GGUF:Q8_0 --rpc node2:50052,node3:50052,node4:50052 -c 8192 # run rpc-server on each other node |
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 2026-10-19.
A project by OnDevice Labs · llms.txt