Select runtimes to compare side by side. Click chips below to toggle selection.
| Metric | Python (CPython) | containerd | CUDA Runtime | AWS Lambda | llama.cpp | ROCm | CTransformers | KoboldCpp |
|---|---|---|---|---|---|---|---|---|
| Score | B71 | B-66 | B-65 | B-65 | C+63 | C56 | D46 | F38 |
| Type | Language | Container | Serverless | |||||
| Execution | interpreted | hybrid | aot | hybrid | aot | aot | hybrid | hybrid |
| Interface | cli | platform | sdk | platform | cli | sdk | sdk | gui |
| Cold Start | 50ms | 100ms | 100ms | 200ms | 100ms | 200ms | 800ms | 1500ms |
| Memory | 15MB | 20MB | 500MB | 128MB | 50MB | 600MB | 200MB | 400MB |
| Startup | 10ms | 20ms | 50ms | 100ms | 10ms | 100ms | 100ms | 300ms |
| Isolation | process | container | hardware | microvm | process | hardware | process | process |
| Maturity | production | production | production | production | production | stable | stable | stable |
| Languages | Python | Any | C, C++, Python | JavaScript, TypeScript, Python, Java, Go, Ruby, .NET, Rust | C, C++ | C, C++, Python | Python, C++ | C++, Python |
| License | Other | Apache-2.0 | Proprietary | Proprietary | MIT | MIT | MIT | AGPL-3.0 |
| Links |