Select runtimes to compare side by side. Click chips below to toggle selection.
| Metric | Python (CPython) | CUDA Runtime | ExLlamaV2 |
|---|---|---|---|
| Score | B71 | B-65 | D47 |
| Type | Language | ||
| Execution | interpreted | aot | aot |
| Interface | cli | sdk | sdk |
| Cold Start | 50ms | 100ms | 1000ms |
| Memory | 15MB | 500MB | 300MB |
| Startup | 10ms | 50ms | 200ms |
| Isolation | process | hardware | process |
| Maturity | production | production | stable |
| Languages | Python | C, C++, Python | Python, C++, CUDA |
| License | Other | Proprietary | MIT |
| Links |