Select runtimes to compare side by side. Click chips below to toggle selection.
| Metric | Python (CPython) | Node.js | Vulkan | llama.cpp | ONNX Runtime | MLX | KoboldCpp |
|---|---|---|---|---|---|---|---|
| Score | B71 | B-65 | C+64 | C+63 | C-50 | D47 | F38 |
| Type | Language | Language | |||||
| Execution | interpreted | jit | aot | aot | hybrid | jit | hybrid |
| Interface | cli | cli | sdk | cli | sdk | sdk | gui |
| Cold Start | 50ms | 50ms | 100ms | 100ms | 500ms | 500ms | 1500ms |
| Memory | 15MB | 40MB | 200MB | 50MB | 300MB | 200MB | 400MB |
| Startup | 10ms | 20ms | 30ms | 10ms | 100ms | 100ms | 300ms |
| Isolation | process | process | hardware | process | process | process | process |
| Maturity | production | production | production | production | production | stable | stable |
| Languages | Python | JavaScript, TypeScript | C, C++ | C, C++ | Python, C++, C#, Java | Python, C++, Swift | C++, Python |
| License | Other | MIT | Apache-2.0 | MIT | MIT | MIT | AGPL-3.0 |
| Links |