← CanIRunLocal · All hardware · All models
OpenBMB · 2.52B parameters · chat, coding, reasoning.
Ollama weight files: Q4_K_M 1.56 GB (hf.co/openbmb/MiniCPM5-2B-GGUF:Q4_K_M, size from the Hugging Face file listing); Q8_0 2.68 GB (hf.co/openbmb/MiniCPM5-2B-GGUF:Q8_0, size from the Hugging Face file listing); F16 5.04 GB (hf.co/openbmb/MiniCPM5-2B-GGUF:F16, size from the Hugging Face file listing).
Yes on 85 machines in this list. The smallest is M1 8GB: Q8_0, about 3.4 GB, roughly 20 tokens per second, via Ollama. ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:Q8_0
ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16lms get openbmb/MiniCPM5-2B-GGUF@Q8_0llama-server -hf openbmb/MiniCPM5-2B-GGUF:Q8_0 -c 8192vllm serve openbmb/MiniCPM5-2B --max-model-len 8192 --gpu-memory-utilization 0.9| Hardware | Runtime | Quant | Memory | Fit | Speed | Command |
|---|---|---|---|---|---|---|
| M1 8GB | Ollama | Q8_0 | 3.4 GB | fits in fast memory | ~20 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:Q8_0 |
| M2 8GB | Ollama | Q8_0 | 3.4 GB | fits in fast memory | ~30 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:Q8_0 |
| M3 8GB | Ollama | Q8_0 | 3.4 GB | fits in fast memory | ~30 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:Q8_0 |
| RTX 3060 8GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~29 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 3060 Ti 8GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~53 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 3070 8GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~53 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 4060 8GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~32 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 4060 Ti 8GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~34 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RX 7600 8GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~34 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 3080 10GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~90 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M1 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~11 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M1 Pro 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~32 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~16 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Pro 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~32 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~16 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~19 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 3060 12GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~43 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 3080 12GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~109 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 4070 12GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~60 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 5070 12GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~80 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 Pro 18GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~24 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 4060 Ti 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~34 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 4070 Ti Super 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~80 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 4080 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~85 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 5080 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~114 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RX 7800 XT 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~74 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RX 9070 XT 16GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~76 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 24GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~16 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 24GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~16 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 24GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~19 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 Pro 24GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~43 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 3090 24GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~111 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 4090 24GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~120 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RX 7900 XTX 24GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~114 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M1 Max (24c GPU) 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M1 Max (32c GPU) 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M1 Pro 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~32 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Max (30c GPU) 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Max (38c GPU) 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Pro 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~32 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~19 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 Max (30c GPU) 36GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~48 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 Pro 36GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~24 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 Max (32c GPU) 36GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~65 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M4 Max (32c GPU) 36GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~65 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Max (32c GPU) 36GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~73 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| RTX 5090 32GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~213 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 Max (40c GPU) 48GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 Max (40c GPU) 48GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~87 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 Pro 48GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~43 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M4 Max (40c GPU) 48GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~87 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Max (40c GPU) 48GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~97 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M1 Max (32c GPU) 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Max (30c GPU) 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Max (38c GPU) 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Ultra 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~127 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 Max (40c GPU) 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 Max (40c GPU) 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~87 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 Pro 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~43 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M4 Max (40c GPU) 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~87 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Max (40c GPU) 64GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~97 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Max (38c GPU) 96GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 Max (30c GPU) 96GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~48 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M3 Ultra (60c GPU) 96GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~130 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M3 Ultra (80c GPU) 96GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~130 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Ultra (64c GPU) 96GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~190 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Ultra (80c GPU) 96GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~190 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Ultra 128GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~127 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M3 Max (40c GPU) 128GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~63 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M4 Max (40c GPU) 128GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~87 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M4 Max (40c GPU) 128GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~87 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Max (40c GPU) 128GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~97 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| DGX Spark 128GB (NVIDIA, ASUS, Dell, Lenovo, HP, MSI, Acer, Gigabyte) | Ollama | F16 | 5.8 GB | fits in fast memory | ~41 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| M2 Ultra 192GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~127 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| 2x Mac Studio M5 Max (40c GPU) 128GB | llama.cpp | Q8_0 GGUF | 3.4 GB | fits in fast memory | ~156 tok/s | llama-server -hf openbmb/MiniCPM5-2B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| Mac Studio M3 Ultra (60c GPU) 256GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~130 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M3 Ultra (80c GPU) 256GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~130 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Ultra (64c GPU) 256GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~190 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| Mac Studio M5 Ultra (80c GPU) 256GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~190 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| 2x DGX Spark 128GB | llama.cpp | Q8_0 GGUF | 3.4 GB | fits in fast memory | ~65 tok/s | llama-server -hf openbmb/MiniCPM5-2B-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 | 3.4 GB | fits in fast memory | ~137 tok/s | llama-server -hf openbmb/MiniCPM5-2B-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 | 3.4 GB | fits in fast memory | ~304 tok/s | llama-server -hf openbmb/MiniCPM5-2B-GGUF:Q8_0 --rpc node2:50052 -c 8192 # run rpc-server on each other node |
| Mac Studio M3 Ultra (80c GPU) 512GB | Ollama | F16 | 5.8 GB | fits in fast memory | ~130 tok/s | ollama run hf.co/openbmb/MiniCPM5-2B-GGUF:F16 |
| 4x DGX Spark 128GB | llama.cpp | Q8_0 GGUF | 3.4 GB | fits in fast memory | ~57 tok/s | llama-server -hf openbmb/MiniCPM5-2B-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 | 3.4 GB | fits in fast memory | ~269 tok/s | llama-server -hf openbmb/MiniCPM5-2B-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 2026-10-20.
A project by OnDevice Labs · llms.txt