Select runtimes to compare side by side. Click chips below to toggle selection.
| Metric | Python (CPython) | Fastly Compute | CUDA Runtime | Node.js | llama.cpp |
|---|---|---|---|---|---|
| Score | B71 | B-69 | B-65 | B-65 | C+63 |
| Type | Language | Edge | Language | ||
| Execution | interpreted | aot | aot | jit | aot |
| Interface | cli | cli | sdk | cli | cli |
| Cold Start | 50ms | 50ms | 100ms | 50ms | 100ms |
| Memory | 15MB | 128MB | 500MB | 40MB | 50MB |
| Startup | 10ms | 10ms | 50ms | 20ms | 10ms |
| Isolation | process | container | hardware | process | process |
| Maturity | production | production | production | production | production |
| Languages | Python | Rust, JavaScript, AssemblyScript, Go | C, C++, Python | JavaScript, TypeScript | C, C++ |
| License | Other | Apache-2.0 | Proprietary | MIT | MIT |
| Links |