Select runtimes to compare side by side. Click chips below to toggle selection.
| Metric | Metal | Python (CPython) | llama.cpp | Candle |
|---|---|---|---|---|
| Score | B74 | B71 | C+63 | D48 |
| Type | Language | |||
| Execution | aot | interpreted | aot | jit |
| Interface | sdk | cli | cli | sdk |
| Cold Start | 50ms | 50ms | 100ms | 300ms |
| Memory | 100MB | 15MB | 50MB | 200MB |
| Startup | 20ms | 10ms | 10ms | 50ms |
| Isolation | hardware | process | process | process |
| Maturity | production | production | production | stable |
| Languages | Swift, Objective-C, C++ | Python | C, C++ | Rust |
| License | Proprietary | Other | MIT | Apache-2.0 |
| Links |