I will put in my two cents on the whole token usage. I use Astro Medium for the most part because that's what I used on Soul. I will obviously admit that Astro does burn tokens…

I will put in my two cents on the whole token usage. I use Astro Medium for the most part because that's what I used on Soul. I will obviously admit that Astro does burn tokens faster than 5.5 Soul, which should be obvious because it's a much better model. Plus the other one had time to optimize over time so it's to be expected that there's going to be a difference in how much it costs to run. With that being said I do feel it's pretty fast. It does burn through it pretty fast. I ran out of usage last night and I just started this morning with almost 100%. I've been running about 5 threads for about 2 hours and it's down to 72%. It's not horrible but there's definitely a huge difference between the previous model and this one. I'm sure they'll optimize it more over time. I don't necessarily think that's a bug. I just think it costs more to run but you can still do a lot with it. One of the biggest things it seems to drain is when you're dealing with 3D, like if it's using Meshy and Blender and multiple things. Even if it's the same amount of time, it seems to use a lot more tokens because it's thinking a lot more, at least from what I can tell. Just my two cents. It's definitely not perfect and it could use some optimizations over time but for me it's kind of to be expected when you're dealing with a more expensive model. The tradeoff is worth it. This is the first time that I've run through it this fast but eventually they'll optimize it.