{"model":{"id":"qwen3.5-35b-a3b-q4","name":"Qwen3.5 35B-A3B Instruct","quantization":"Q4_K_M"},"hardware":{"slug":"rtx-5090","name":"NVIDIA GeForce RTX 5090","memoryGb":32},"verdict":"yes","localVerdict":"local_feasible","fitLevel":"Excellent","estimatedLoadGb":20,"usableMemoryGb":28.8,"estimatedTokensPerSec":117.5,"kvCacheByContext":[{"contextTokens":8192,"kvGb":0.16,"totalGb":20.16,"fits":true},{"contextTokens":16384,"kvGb":0.31,"totalGb":20.31,"fits":true},{"contextTokens":32768,"kvGb":0.63,"totalGb":20.63,"fits":true},{"contextTokens":65536,"kvGb":1.25,"totalGb":21.25,"fits":true},{"contextTokens":131072,"kvGb":2.5,"totalGb":22.5,"fits":true}],"upgradeGpu":null,"ollamaCommand":"ollama run qwen3.5:35b-a3b","reportUrl":"https://modelfit.io/can-i-run/qwen3.5-35b-a3b-q4-on-rtx-5090/","source":"https://modelfit.io/","license":"CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)","caveat":"Estimates from the ModelFit engine, not measurements."}