Get startedLeaderboardDecode calculatorModelsReportsHardwareBenchmarksMarketplaceRentalsProAPI Docs
Language
Total runs
Highest
Best memory ceiling
Median
Lowest
nvidia_Nemotron-Cascade-8B-GGUF — 62.4 tok/s on RTX 3060 · 12 GB ×2 | Localmaxxing