4 ms·

You can use a lower-end GPU (like the RTX 3060), which also uses less energy. But you are right, you won't be encountering model API costs when running it loca

You can use a lower-end GPU (like the RTX 3060), which also uses less energy. But you are right, you won't be encountering model API costs when running it locally.