Select runtimes to compare side by side. Click chips below to toggle selection.
| Metric | CUDA Runtime | llama.cpp | ROCm | LLM (Python CLI) |
|---|---|---|---|---|
| Score | B-65 | C+63 | C56 | D49 |
| Type | ||||
| Execution | aot | aot | aot | hybrid |
| Interface | sdk | cli | sdk | cli |
| Cold Start | 100ms | 100ms | 200ms | 500ms |
| Memory | 500MB | 50MB | 600MB | 100MB |
| Startup | 50ms | 10ms | 100ms | 100ms |
| Isolation | hardware | process | hardware | process |
| Maturity | production | production | stable | stable |
| Languages | C, C++, Python | C, C++ | C, C++, Python | Python |
| License | Proprietary | MIT | MIT | Apache-2.0 |
| Links |