Select runtimes to compare side by side. Click chips below to toggle selection.
| Metric | CUDA Runtime | llama.cpp | CTransformers | KoboldCpp |
|---|---|---|---|---|
| Score | B-65 | C+63 | D46 | F38 |
| Type | ||||
| Execution | aot | aot | hybrid | hybrid |
| Interface | sdk | cli | sdk | gui |
| Cold Start | 100ms | 100ms | 800ms | 1500ms |
| Memory | 500MB | 50MB | 200MB | 400MB |
| Startup | 50ms | 10ms | 100ms | 300ms |
| Isolation | hardware | process | process | process |
| Maturity | production | production | stable | stable |
| Languages | C, C++, Python | C, C++ | Python, C++ | C++, Python |
| License | Proprietary | MIT | MIT | AGPL-3.0 |
| Links |