{"model":{"id":"gemma4-12b-q8","name":"Gemma 4 12B (Q8)","quantization":"Q8_0"},"hardware":{"slug":"rtx-4070","name":"NVIDIA GeForce RTX 4070","memoryGb":12},"verdict":"tight","localVerdict":"local_slow","fitLevel":"Heavy","estimatedLoadGb":12.8,"usableMemoryGb":10.8,"estimatedTokensPerSec":4.6,"kvCacheByContext":[{"contextTokens":8192,"kvGb":1.5,"totalGb":14.3,"fits":false},{"contextTokens":16384,"kvGb":3,"totalGb":15.8,"fits":false},{"contextTokens":32768,"kvGb":6,"totalGb":18.8,"fits":false},{"contextTokens":65536,"kvGb":12,"totalGb":24.8,"fits":false},{"contextTokens":131072,"kvGb":24,"totalGb":36.8,"fits":false}],"upgradeGpu":{"slug":"rx-7900-xt","name":"AMD Radeon RX 7900 XT"},"ollamaCommand":"ollama run gemma4:12b-it-q8_0","reportUrl":"https://modelfit.io/can-i-run/gemma4-12b-q8-on-rtx-4070/","source":"https://modelfit.io/","license":"CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)","caveat":"Estimates from the ModelFit engine, not measurements."}