Maybe that was true with Fable as well. But one prompt which took 5 minutes of clock time, used 50% of the resources of a Pro account for the 5 hour window.
I used 50% in just one task(3d gamedev) on AI Pro in Antigravity desktop. It is not my daily driver anyways, but it was nice to see that it is actually performing and completing the task.
Automated model routing is just immature right now. Given how fast things are moving, though, I'd expect these resource allocation pitfalls to be ironed out soon.
I agree, I have tried it for making motion graphics using Argon, in middle of code generation maximum tokens got used and auto switched to 3.1 pro
At least it picked that, mine goes to flash lite
At some point human labor will be competing with AI simply by charging less than the models charge per hour for the same task
Argon is a 4$in/20$out model; how far do you expect a 20 dollar a month plan to subsidize?
Wait until you hear token pricing for Claude.
That's in the Opus 5.5 range, no?
Get yourself an OpenCode Go sub for $10/month, it will go much much further